Document Type: 
Document Number (FOIA) /ESDN (CREST): 
Release Decision: 
Original Classification: 
Document Page Count: 
Document Creation Date: 
November 1, 2016
Sequence Number: 
Case Number: 
Content Type: 
PDF icon CIA-RDP82-00850R000500060015-4.pdf10.93 MB
APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500060015-4 FOR OFFICIAL USE ONLY JPRS L/10502 ~Aay 198~ ~ . - U~SR Re ort p . CYBERNETICS, CONiPUTERS AND ~ AUTOMATION TECHNC~LOGY CFOUO 7/82) ~XCERPTS FROM 'COMPU~ERS A~1D COMPUTER NETIMC~RKS' ~ FB~~ FOREIGN BROADCAST INFQRMATIOR! SERV!^E FOR 6F'F'ICUL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/42/09: CIA-RDP82-40850R000500460015-4 NOT~ JPRS publications contain information primarilq from foreign = newspapers, periodicals and books, but also from news agency transmisaions and broadcasts. Materials from foreign-language so+urces are translated; those from English-language sources are trar.scribed or reprinted., with the original phrasing and other charact~ristics retain~:d. Aeadli~nes, editorial reports, and material enclosed in brackets are supplied by JPRS. Processing indicatars such as [T xt] ~ or [Excerpt] in the first line of each item, or following . last line of a brief, indicate how thE originaY information was ~ proces~ed. Where no processing indicator is given, the infor- mation was summarized or extracted. Onfamiliar names rendered phonetically or transliterated ara enclesed in parentheses. Words or names preceded by a ques- tion mark and enclose~ in parentheses were not clear in the - original but have been supplied as appropriate in context. - Other unattribvted parenthetical notes within the body of an item originate with the source. Times within 3.t~s are as given by source. The ront~ents of this publicatiQn in no way represent the poli- cies, views or at.tit~~des of the U.S. Government. ~ COPYRI(~IT I.AWS AND REGULATIONS GOVERNING OWNERSHIP OF y Ml~TERIALS REPRODUCED HEREIN REQUIRE THAT DISSEMINATION OF Tfil~ PUBLI~.~TION BE RESTRICTED FOR OFFICIAL USE ONLY. APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~ FOR OFFICIAL USE ONLY JPRS L/10502 7 May 1982 USSR REPORT , - CYBERNETICS, COMPUTERS AND AUTOMATION TECHNOLOGY (~OUO 7/$2~ EXCERPTS FROM 'COMPUTERS AN~ COMPUTER NETWORKS' CONTENTS Annotation 1 = Contents 1 _ Information Transmission Equipment 3 YeS Remote Data Transmi.ssion Sys~;sm 12 Yes Computer Specifications .27 _ ASVT Computers 32 . Stor~ge Devices 3Lt ; Multiprocessor and Multimachine Computer Systems 43 rt-222 52 riir~i~nax ana sunma 53 vx-ioio 55 ~ Design Principles of Multiprocessor Computer Systems 58 E1'brus 6L~ Informatio~ Coding and Data Preparation 69 ' KOI-7, KOI-8 and DKOI Codes. 71 Punch Cards ?5 _ - a- [III - USSR - 21C S&T FOUO] FOR OFF[C[A L 't,$L ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2447/02/09: CIA-RDP82-00850R000500464415-4 F'OR OF'FICIAL USE ONLY _ I Paper and Magnetic Tapes . ?9 Data Preparation and Conversion ~2Y].C@3 83 ' Central Processors 89 Microprog~camning 103 Network Principle of Control 106 . Interrupts ...........e 108 , YeS Computer Arithmetic-Logic IInits 116 ; Peripheral Devices 123 . Input Devi.ces 126 Output Devices 135 : = Operator-Computer Co~nunicati ms Devic~es 145 I External Storage Dev~~ces 151 AP User Stations and Remo~te Data Processing 162 ~ Bibliograph3r 171~ -b- FOR Ot~i~IAL iJSB ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OFF'ICIAL USE: ONLY COMPUTERS AND COMPUTER NETWORKS Moscow EVM I VSiCAISLITBI.'NYYE SETI in Russian 1980 (signed to pre3s 20 May 80) pp 2-3, 283-303, 326-327 [Excerpts from book "Computers and Computer Networlcs", 'oq Vasiliy Nikolayevich Kriushin, Inna Nikolayevna Buravtseva, Nina Mikhaqlovna Pnshkina and Nina Grigor'yevaa Chernyak, Izdatel'stvo "Statistika", 18,000 copies, 328 pages] [Excerpts] Annotation. The functional principles of compnters and systems ~ are outlined. The maCer3.a1 is based on third-geaeration ccnnputers. Basic attention is dewted to loqic orqanization of computers and systems, the � character.istic features of encoc]ing economic data, the structure of peripheral ~ equipment, the characteristics of individual devices and their operating capa- bilities. The principles of designinq collectiv~e-use oomputer netwcrks, the eq_uipment included in the netvrorks and user interaction with the computer are considered. � The material in the textbook corresponds to the curriculum of the course "The Confiqurat~.on of Computers and O~niputer Netw~rks" for students at higher edu- cational institutions studyinq specialty No 1738. The book will be of inter- est to students and spezialists workinq in the field of campu~er applications in automated control systems asd at computer centers. ~ Introduction . ~ At the preseat time there are roughlp 3,000 computer centers in the country [3]. More than 3,000 automr-.,ed control centers were in use at the be- qinninq af 1977 [2]. In 1978 alone more than 400 auton~ted control syster~~ for accountinq, planninq and manaqement were set up [32]. More than 300,C~d specialista in the country are ~aorkinq on development and operation of auto- mated control systems [33]. Contents Page Annotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Section 1. Information-Logic Fundamentals of Computers ~ ~Chapter l. Ge~eral Information on Machine C7onversivn or Data. 11 1.1. N~anber system used in computers . . . . . . . . . . . . . . 11 _ 1.2. Arithmetic operations on numbers . . . . . . . . . . . . . . 15 , ~ FOR OFF'ICIAL USE ONtY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 MVK lsrrlt,lAL. U~~. VIVLY Chagter 2. Inforn~ation Codinq aad Data Preparation . . . . . . . . . . 24 2.1. Methods of encoding economic data �or machine pmcessinq. � 30 2.2. Information carriezs . . . . . . . . . . . . . . . . . . . 2,3. Data preparation and information rewritinq devices from 39 one carrier t~o another . . . . . . . . . . . . . . . . . . . Chapter 3. Nachine Display of Data and Control Information 45 3.1. Machine display of data . . . . . . . . . . . . . . . . . 45 3.2. Display of control information . . . . . . . . . . . . . . . 48 Section 2. Computer Configuration Chapt~r 4. General Inforn?ation on Computer Cvnfiguration ~2 4.1. The conoept of the computer system and its aonfiquration. . 52 - 4.2. The families of third-qeneration computers. . . . . . . . . 58 4.3. Composition and basic characteristics of the Unified 6~ Camputer System and m~dular comPuter equipment. 68 4.4. Basic concepts on software ornaposition. . . . . . . . . . . Chapter 5. The Central Processor . . . . . . . . . . . . . . . . . . . 72 5.1. Block diaqram of the processor . . . . . . . . . . . . . . 72 - 5.2. The central control device . . . . . . . . . . . . . . . . . 76 5.3. Microproqram and network px~nciples of computer ~.ntisool 86 5.4. Multipmqram operatinq mcde and interrupt system 40 5.5. Designatic~~ and composition of arithmetic~leyic device 99 5.6. Performance of binary arithmetic operations 104 5.7. Performance of decimal aritlnaetic operations. 112 5.8. Performance of logic operations . . . . . . . . . . . . . . 115 115 - Chapter 6. Computer Storage D~vices . . � � � � � � � ' ' ' ' ~ ~ ~ ~ 115 6.1. Classification and basic parameters of storaqe devices. 125 6.2. Internal Storaqe . . . . . . . . . . . . . . . . . . . . . 6.3. Systems of ineuiory organization and methods of ine~arY 134 protec~ian . . . . . . . . . . . . . . . . . . . . . . . . 6.4. High-sp~ed memory . . . . . . . . . . . . . . . . . . . . . 145 148 6.5. Read-only storaqe devices . . . . . . � � � � � � � � � 155 Chapter 7. Input-Output Chan~nels . . . . . . . . � . . � . . . � . . ' 155 7.1. General informati~~~n . . . � � � � � � ' ' ' ' ' ~ ~ ~ ~ ~ 161 7.2. Orqanization ~i ir.~.put-output. . � � � � � � ' ' � ` ~ ~ ~ ~ 166 ~ 7.3. The multiplex char;.hel . . . . . . . . . . . . . . . . . . . 7.4. The selector chanael. . . � � � � � � � ' ' ' ~ ~ ~ ~ ~ ~ ~ 180 7.5. The input-c+utput interface . . . . � � � � � � � � � � ' 185 Chapter 8. Peripheral Devices . . . . . . . . . � � � � � ' ' ' ~ ~ ~ 185 8.1. Desiqnation and classification of peripheral devices. 188 8.2. Conaputer input devices . . . . . . . - � � � � � � � � � � g.3. Computer output~ devices . . . . . . . . . . . . . . . . . . 198 8.4. Operator-oomputeX communication devices . . . . . . . . . . 207 5.5. External storaaA devices . . . . . . . . . . . . . . . . . 213 2 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R000540060015-4 FOR OFF7CIAL USE ONLY , Chapter 9. Mu~,tiprocessor and Multimachine Cc?mputer Systems 224 9.1. General infdrmation on multiprocec~sor and multimachine oomputer systeu~s . . . . . . . . . . . . . . . . . . . . . 224 9.2. Desiqn principles of multimachine oomputer systems (oomplexes ) . . . . . . . . . . . . . . . . . . . . . . . . 2 31 9.3. Design principles of multiprocessor computer systems ~ (oomplexes ) . . . . . . . . . . . . . . . . . . . . . . . . 240 Section 3. Computer Netw~orks Chapter 10. General Information on Computer Netw~orks 251 10.1. Designation and basic principles of developmen~. 251 10.2. Classification . . . . . . . . . . . . . . . . . . . . . 257 10.3. Basic functional principles . . . . . . . . . . . . . . . 262 Chapter 11. Con~unications ~ystems . . . . . . . . . . . . . s . . 267 11.1. Basic characteristics of coamunicat~ons channels 267 11.2. Meth:~ds of error protection . . . . . . . . . . . . . . . . 274 11.3. Information transmission equipment . . . . . . . . . . . . 283 ~ Chapter 12. Remote Data Processing Sys~tems . . . � � � � � � � � � � � 293 12.1. General information� � � . � � � � � � . . � � � � � � � � 293 12.2. Remote data processinq hardware . . . . . . . . . . . . . 295 Chapter 13. User Stmtion Systen~s . . . . . . . . . . . . . . . . . . . 303 - 13.1. General information . . . . . . . . . . . . . . . . . . . 303 13.2. User stations of the YeS renpte data processing system 307 Chapter 14. Foreign Computer Netw~rks . . . . . . . . . . . . . . . . 314 Biblioqraphy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 List of Abbreviations Used in Text and Illustrations. 3~1 Chapter 11. Communications Syste~as 11.3. Infonnation Transmission Equipment The enornnous importance of conununications equipment was pointed out by V. i. Lenin, who wrote in 1918 that "socialism without the mails, telegraph and ma- chinery is an empty phrase."* Selecting the means of coumunication depends on a large nwnber of factors: the tzwmber of users, the speed and reliability of transmission, the length of the communications channels and so on. T~legraphic conmaunications equipment is most efficient in thcse cases which require high reliability of transmission, telephone conmunication equipment is preferable for higher speed and facsimile co~aunication equipment is best for "transmi.tting drawinqs and figures. *"Polnoye sobraniye sochineniy" [Compl~ste Works], Vol 27, p 278. 3 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OFF7CI,e1L USE ONLY The basic equipment used in teleqraphy coa~rises various types of teleqraph operation. The con~temporary models use the modified international teleqraph code MTK-2 recommended for use by the International Teleqraph and Telephone Consultative Committee. Tl3is code is a unifortn, five-element code in which zero is transmitted by arn absence of current and one is transmitted by the presence of current. For this reason all electrical com~onents are divided accordingly into currentl.ess atid current components. Each oombination transmit~ed by teleqraph channel in MTR-2 code has a length of Sto, where to is the length of a single signal. Hawever, when telegraph equipment operates in the start-stop mode, two additional service ~mponents are introduced. The first one, the start, beinq without current, is transmi~- ted before the code combinntion and also has a length of to. The sevond com- ponent, the stop, which completes transmission of the code co~abination and ~has the current position "1," is longe~-1.5-2.0 to. Thus, the total lE::~~h of one code combination transmitted by telegraph channels in N~K-2 code is 7.5 to. _ In addition to the general characteristics described above, various other in- dicators are used to define telegraph operations: correcting capability, which characterizes the quality of vaork of the teleqraph apparatus and is determined by the magnitude of maximum marginal distortion w3th which it is still possible to record an error-free ~ignal. = This capability may be theoretical, effective (~hat is, measured under real operatinq caazditions) and nominal (which is the lowest for a group of similar devices); theoretical probability, equal to the maxim~nn nianber of w~ords which can be trarrsmitted and received by the teleqraph apparatus in aa hour: technical productivity, which takes into account the time of transmitting useful information and servi~e infoxmation= operatinq pr~ductivity, which reflects the design characteristics of the , equipment, operator qualifications and so on. All telegragh equipment is basically similar in design and has the following primary functional parts: the transmi~ter, which enters and converts the information to be transmit- ted. It vomprises a keyboard, a coding device with five steel bars, each of which may occupy one of two fixed positions correspondinq to "1," and "0," by which the coding device carries out the coding, the transmitting distributor, - ~ whose primary purPose is to convert parallel code combinations formed by the bars irito a sequence of electrical sig~nals. Z'he xotational frequency of ~he distributors, which is a fixed value in all telegraphs, determines the len3th of a single signal to; the receiver, whose functions are to receiv~e, convert and print the in- formation received from communications channels on papertape. The receiver consists of a distributor which percforms the inverse conversions relative to the transmission, the decoder with five decoding bars and a printer= FOR OFI~'IC'41AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FOR OFF7CIAL USE ONLY a drive mechanism= auxiliaYy assewblies and dv~vices, includinq mechanisms to feed the paper and inkinq tapes, the reqister switch mechanis~a, a self-locking device and so ~on. Teleqraph lines use the domestically produced ST~35, ST-2M, STA-2M, RTA-6, STA-M67, RTA-7 and other teleqraph equipment p 1 us the East Germnn Z"-51 aad T-o: units. The STA-2M telegraph, which is a modification of the ST-35, uses I~l1'R-2 oode. It has a speed of 50 bauds and a maximum operatinq ranqe of 550 ki].ometers. The letter "A" in the STA-2M telegraph means "at~tomated" because transmitter and reperforator attact~ments are connected to it, which permits it to be used for transmitting a prepunched tape and for rec~iving informa+tion fran the chan- nels on both papertape and punch tape. The availability of.these attachments leads to full use of com~aunications chan- nels and the capability of preparing punched tapes in advance malces it possi- ble to improve the quality. Moreover, the use of punched tapes permits automa- tion of the message relay operations at central telegraph stations and makes it possible to use punch tapes as information carriers for computers. The STA-M67 t~legraph differs a~ly sliqhtly from the STA-2M. The page-printinq start-stop teleqraphs were a further development of tele- graph technology. The RTA-C paqe-printing teleqraph with transmitter and re- perforator operates at a speed of 50 and 75 bauds, has a roll of paper 215 millimeters wide and can receive three oopies of text simultaneously. " The RTA-7 page-printinq telegraph, unlike the electromechanical devices oon- sidered abov~e, is classified as an electromechanical teleqraph in which cer- tain mechanical assemblies such as the distributor, coder, decoder and a few others are replaced by electronic devices. 't'he RTA-7 operates at speeds of 75 and 100 bauds and has three reqisters, automated attachments, a noise sup- pression filter, a radio receiver and a number of other auxiliary units. _ Telegraph communication can be orqanized usinq either qeneral-purpose eoiomuni- cations included in the system of USSR Ministry of Communications and serving various institutions and the public through a netw~ork of communications depart- ments or by means of user teleqraphy, which serves those instituti,ons that have telegraph equipment connected to the user telegraph station. One of the varieties of user teleqraph equipment is the international Telex telegraph, designed for transmitting messages to other countries. Facsi4nile oonmunications apparatus is used to transm~t fixecl images (docunaents, photegraphs, drawings, tables, textual material and a~o on) through oo~nmunica- tions channels. According to the definitions of the International Telegraph and Telephone Consultative Committee, the terms "phototelegraph" and "phototel- egraphic apparatus" should refer only to the equipment used to transmit and - FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 rvec vrr~..~na. ~~c. .~i~L~ . receive half-tone images. The more general ternu is "facsimile apparatus,"* related to devices that transmit both half-tone and facsimile originals. Zhere is a fairly broad ranqe of facsimile oo3nnunications equipment including the followinq devices: FTA-PM, Aragvi and Shtrikh, which are used only for receiving and trans- mitting facsimile images measuring 220 X 150 millimeters (220 X 300 mi.llimeters in the FTA-PM), they ha~re transmission times of 12.5, 6 and 2.1 minutes, re- spectively, and the imaqe is printed on ordinary papert i,adoga, designed to transmit hydrometearological charts of unli.mi.ted length and up to 480 millimeters wide. A sheet measurinq 480 X 690 milliaaeters requires 22 minutes for transmission= Neva is used for working with both hal~-tone and facsimile originals. It requires 6 minutes to transmit an original measuring 220 X 300 millimeters; Gazet~l and Gazeta-2 are used to transmit newspapers= fox this reason ' they are larger than the original (420 X 610 millimeters), can receive copies on photographic ~ilm and have a more riqid requirement for misalignment of imaqes--not m~re than 1 millimeter for each 100 millfaneters of the page. The devices can transmit one newspaper page in 50 and 2.3 minutes, respec+:ively. All facsimile devices consist of transmitting and receiving parts. Z'he trans- mittinq part includes: a scanning device for breaking the image of the oriq- inal down into elements, a light-optical system by means of which a light beam :s naved across the surface of the original secured to the surface or a rewlvinq drum (as a function of planar cr cylindrical scanning), a photoelec- tric conrerter that converts the light reflector from the oriqinal into elec- tricial pulses whose value depends on the image brightness, a video sfgnal oonverter, phase and synchronization devices and so on. The basic devices of the receiving part of the facsimile apparatus are the video signal converter, the recording and scanning devices and the phase and synchronization units. They convert electrical pulses fed from the communic~~- tions channel into a luminous flux which is then projected onto the surface of phctoqraphic paper (this method is called the closed method because the image beoomes visible only after the photographic paper is properly processed? or on- to electrochemical, electrothermal and other types of paper (this method is called open because the copy appears inmediately upon receipt). Among the traditional means of conmunication are varieties of telephone equip- ment, including the TA-60, TA-65 and TA-72 (Soviet Uaion), TsB 621/65 and TsB 631/65 (Poland), 63 St (East Germany), TA-66a (Czechoslovakia) and others, the ATS-47, ATS-54 and ATSK automatic telephone exchanges and so on. *A facsimile is an e~ct reproduction at the receiving station of the image beinq transmitted. 6 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500460015-4 Among the various means of canmunication, the most attention is devoted to building and using data transmi.ssion equipment (APD). This usually includes an error protection device (OPZ), designed to enter verification symbols to protect the data being transmi.tted against errors, a memory unit whose ~apac- ity must be sufficient to store at least one code combination, a control de- vice which provide for interaction among all assemblies of the data transmis- - sion equipment, input-output devices and a signal conversion unit (UPS). The signal canversion unit is used because the data to be processed by computer are written in binary form and represented by square-wave pulses. No addition- al operations are ordinarily required to transmit them by teleqraph channels. But the data must be agpropriately converted to transmit them by telephone channels, which is done by using a sc-called carrier frequency selected in the ' middle of the bandpass and the data are entered in i~. This process, called modulation, greatly increases transmission speed. Modulation can be frequency (the frequency of the carrier oscillation is subject to variation), ampl:tude or phase. Modulation is accomplished in special devices called modems whir.h generate the needed sitie curve and modify it according to data received. Demodulation of the carrier frequency is also carried out in modems. Some of the data conversion units such as the Akkord-50, the Akkord-1200 and the Minsk-1500 (Figure 11.4) operate in a semiduplex mode that provides data transmission by telegraph channels at speeds of 50 and 100 bauds. Transmis- sion reliability is 3�10'~. Information is entered from papertape by an FF- 1500 photoreader (Czechoslovakia) and retrieved to papertape by a PY,-150 tape punch that is connected to the Akkord-50 by a remote inteqration, perforation and readinq unit. Inforn?ation is entered and also retrieved by page-printing telegraph. Connections are made using the call device included in the Akkord- 50 set. Cyclic coding and resolving feedback are used to ensure qood-quality transmission. � ~ wa~Mai npa6o (1 e~wQwmi qou~iqo f,r-~saa nn~rso � Y C, ~NMMOpd- � TC/lCi O N171 ~R~r~taJO~-SD' y~ ~ Ka,~6 p~v /1/l-130 � FS-1100 o7A (4) ~ ~q PTq Figure 11.4. Block Diagram of Data Transmission Using the Akkord-50 Data Transmissiv;n Equipment Key : 1. Call ~evice 4. Page-printinq telegraph 2. Signal conversion unit 5. Akkord-50 3. PIr 50 tape punch 6. Telegraph channel 7 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 HUI~ UrH'1(:IAL U,~. UNLY The Akl:ord-1200 data transmissi~n equipment is used for with tele- phone comnunicatiAns channels and functions both in the semiduplex (switched channels) and duplex (seqregated ahannels) modes at speeds of 600 and 1,200 bauds. High reliability of data t~ansmission (no worse than 10-6) is achieved by using cyclic oodes with the polynomial x~~ + x12 + xs + 1 and resolving feedback. The Akkord-1200 operates in the following modes: "telephone," c�lhich is designed for service communications, "data transmission," which only transmits data, "data reception," which only rec~eives data and "internal," which is used to check the working condition of the equipment. The Akkord-1200 set of equipment includes a PL-150 tape punch, aa FS-1500 photoreader, a calling and ringing device and~an Akkord-1200 PP transceiver. If the user station does nothing but transmit data, it will have an Akkord- 1200 PD transmitter operating in three modes: telephone, data transmission and internal, instead of the Akkord-1200 PP. The transmitter has a codinq device that foYms ~he check combination of the cyclic code, a data input oon- trol unit thaic converts parallel coda into series code, writes data into the memory unit and forms the service char~icters of the data unit, a phase device to determine the beginning point of phasing and to establish phasing modes (the synchx+onous method of transmission is used), pulse shaping control arid _ reverse channel signal devices and so on. On the other hand, if the user station only receives data, the ~et af equip- ment includes only an Akkord-12 PM~receiver which operates in the telephone, data receiving and internal modes. The receiver includes a decoding unit that detects errors in the block of dzta received, a unit, an input reg~.ster desxqned to convert data from series to parallel code, a control pulse and re- verse channel signal shaper and other equapment. The Akkord-1200 has a memory unit for simultaneous storage of two data blocks and a modem-1200 that converts digital signals received from terminal equip- ment to frequency-modulated sigraals suitable for transmission by telephone channels and for inverse conversions. The Minsk-1500 automati~ data transmission equipment also transmits by tele- phone channels, but it has the capability of being directly vonnected to ~the Minsk-22 oomputer. An inverse code is used in the equipment to protect against errors; thE reliabili~y of transmi~sion is 10-6. The DFE-550 automatic data transmission equipmQnt (East German) is also used to transmit data by telephon~ channels at speeds of 600 and 1,200 bauds and em- ploys cyclic codes which provide a data transmission reliability of 10~6. The information read from the punched tape is fed to the memory unit, designed to store data until it has been verified at the receiving station, and to the coding unit where the data is conve+ted from parallel to series code, and cy- clic coding operations are performed. The informatior. is ~hen fed to the modem where it is converted so that it can be transmitted by telephone channels. 8 FOR OFFICIAL ~JSE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R000540060015-4 FOR 1 . When infos-mation is received from comm~unications c~+annela, inverse conversion takes piace in the awde~n in order to emit DC p~ulses to the squignent. infori ~?u9tion is converted to parall~l code by a converter that is included in the d~arodinq unit. The infonaatioa is then fed to the memory unit, to that zone in which it was located on the transmitting side. Then the correctness o� the trsnsmission is checked and if so. ~ confirmation siqnal is sent to the trans- mitting party by the feeclba.;~ .hannel. if the transmission was incorrect, a signal to repeat is sent. In view of their differences in information vodinq, speed of input-output from the computer and transmission by cammunications channels, shape of signals and so on, it is necessary to mattch parameters to operate data tx;ansmission and computers together. One of the methods for this assumes the use of papertape, which ensures a fuller load on conmunicatione chanaels because the can be prepared throuqhaut the w~orkinq day, while it is transmitted on1.y mt a strictly determined tin?e. But with this technique instantaneous datn pi+ocess- ing is impossible. Therefore, a different technique of matchinq the caaputer and communications channel is better that requfres the use of special integra- tion devices such as the Minsk-1560. Thfs device can connect up to 32 tele- graph channels with a transmission speed of 50 bauds in each line to a Minsk- 32 or Minsk-23 computer, the use of unswitched telegraph lines (in this case the nwnber of user stations canr~ot exceed 32) and connection by four telephone lines through Akkord-1200 or the Minsk-1500 data transmission equipment. The Minsk-1560 device also locks out all coma~unications lines when there is a malfunction in the computer and any particular line while it is operatinq or in the absence of lines. The Minsk-1560 includes line equipment, the memory unit, a control device, a calling device, a page-printinq teleqraph and so on~ When data is being entered from telegraph channels, it is first fed to the individual line equipment in which the telegraph messages are oonverted into~ pulse signals and then to the mennry unit and shift reqister where the series code is converted to parallel code. The data are then fed through the exchange reqisters to the magnetic core storage of the computer to the address indicated in the appropriate duty register. A region of the maqnetic core storaqe and duty reqister correspond to each of the 32 comnunications lines. When data is being retrieved from the computer, an evenness check is made in the exchange register and it is converted from parallel code to s~ries code in the shift register. The teleqraph messaqes are formed by an electronic relay. All input-output operations are coordinated by a special input-output s~bmu- tine entered in the computer in advance. Besides the above communications equipment, communications systems and computer networks must in practice include various switching devices, from very simple electromechanical to switching centers that use electronic equipment. ~ The first message-switching devices were developed for telegraphy, where a rotating 360-degree selector was used as a switching element to connect users. FOR OFF[CIA.L USE ONLY ~ APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500460015-4 F'OR OMFICIAL U5~ UNLY The brush of the selector in such a commutator moved across a contact field and was oonnected to the user line through a pluq selector. ~ With the invention of the telephone and the spread of telephone communications, manually operated switchboards appeared, followed by automatic switchboards, including step-type and circular selector switchboards. The further develcp- ment of switching equipment led to coordinate selectors and matrix-type relay systems. In recent years semiconductor oomponents have begun to be used in place of relays in the connecting assemblies, which has led to the establish- ment of electronic switching centers. Figure 11.5 presents a block-diagram of a comaaunications switchinq center used in a computer network, which includes the following: a line equipment unit, a input-output unit, a connectinq component unit, storaqe device and a ' control unit. KoMOner 6n~ ~ , ~ p~Yl~ , ~ . ~ 6B$ 3~ 6C3 ~4 3y ~6~ . . ~ KaNane~ ~ � A~u ~ 6/!0 ( 2 ~ y~f" 5 Figure 11.5. Block Diagram of Communications Switchinq Center _ Key: 1. Conanunications channela 4. Connecting vomponent unit 2. Line equipment unit 5. Control unit 3. Input-output unit 6. Storage device Th~ line equipment unit integrates the communications channels by mearis of which r~note users are connected to a spec~fic awitching center. While per- forming this function, the block constantly monitors the status of the com- munications channels so that when a request is received from a user to estab- lish com~?unications, the necessary line can be connected, the signal decoded correctly and communications stopped at the end of the transmission. Each unit cor~responds to a specific line and includes the equipment assigned to that line. Modems are installed in the line equipment unit to convert input and output signals into forms suitable for internal switching at the sw~tching center and for transmission by communications lines. The input-output unit is designed to identify and assign service characters - (start and stop messages, taqs for the beqinning and end of inessages and so on), verify that the equipment is connected carrectly to the beginninq, deteri mine the address of the prececlinq switching center and perform various other functions. 10 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~ The connectinq oomponent unit ca?rries the primary load of establishinq commun- ications between users. '1"liis unit is the mcst in~portant and has the qreatest specific weight at channel switchinq centers. The principle function of the connectinq camponents is to ensure rapid, good communications between input arid output us~r lines. One of the methods of connection is titae-sharinq (sps- tial switchinq, switching physically distinct circuits) where the coordinated structure of the switchinq field is a matrix formed of intersection points be- tween input and output lines. The confiquration of the switchinq field may vary. Thus, the nwqber of conr?ec- tinq assemblies in a switchinq field conatructed on a single-staqe noninter- locking schems (Fiqure 11.6a) is equal to the product of the number of input and output lines. For syatema with up to 1,d00 inputs and outputs, the most widespread system is the three-staqe scheme (Figure 11.6b), khich ensures a, minimum nwnber of connections with low probability of losinq the call. Be~xode~(1) � " ~ � . . . 0 t . e C ( ( 2~xoaa �A Q c eaX~it t 1 b Figure 11.6. Switching Field Structures Key: ~ 1. Outputs 2. Inputs - When desiqning a communications center, an effoxt should be made to minimize the number of points throuqh which the connecting path travels and ensure that the probability of interlocking is insignificant. The storage device may be small memory units in the form of registers (chiefly in the central switchinq channel) designed to store infornation for the time required to process the call and to fornt the connecting chanhel, to store data on bypass routes and to store the various types of tables and references neces- sary to select routes and distribute messaqes. The memory of the switchinq centers may be organized by using large-capacity memory units, which is more typical of inessage-switching centers. In this case several alternatives are 11 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 �'V� V~'l~~.tA~/ VI.I~+ V~\VL possible for usinq the memosy unit. According to one of them, tw~o menwry wlits are installed~-one is located at the input of the messaqe-switchinq center and - the other at its output. Upon receipt, incosainq messaqes are written into the first menwry unit until a sufficient number of them have been collected to or- qanize switchinq or until the end of the message siqr~al is received. A�ter � this they are rewritten into the inpu~ memary units. Output to the communica- tions channel occurs either as the messages oame in on the "first received, first served" principle or according to their priority. Aaother alternative assim?es the existence of one n~emory unit that stores all information. When a signal is received that indicatea the necessity to switch to messaqe processinq (verify the correctness of receivinq, extract service information and so on), the output channel is checked to see ~hether it is occupied. After all these operations are performed and the initial channel is free, the data are fed to the coamnunications channel. This procedure, which is - the n?ost popular today, found application in practice only after the appearance of magnetic tape, dr~aa ar~d disk men~ory units. The control unit coordinates the work of all the units of the switchinq center. Thus, the initial call sent by the user is received by the control unit, which assigns the sequence of performinq the operations by which the line equipment unit will scan and identify the line on which the call came, analyze all pos- sible directions of further messaqe travel upon finding a free path, sead in- structions to the connectinq component unit to make the required oonnection and so on. in computer netw~orks, conanunications processors, concentrators and one or several com~uters can be connected to the switchinq centers. This ~qu~pment is used for partial message processing, route selection and certain other operations. ~ The American IBM 5910 system can be cited as an example of a message-switching center. Up to 480 telegraph channels can be connected to it thsnugh line unit 5974. The system has paired 5978 processors, one of which is in working oon- dition and the other in reserve at any mmnent~of time. The memory unit is composed of paired magnetfc disks. The duplication af equipment included in this system provides hiqh operatinq reliability. With an average message length of 300 characters, the switchinq device processes up to five messaqes per secand. The message-switching center is connected to the computer throuqh a standard IBM selector channel. Chapter 12. YeS Remote Data Transmission System 12.1. General Informatio~ Remote data processinq systems, which are one of the foundations of computer networks, are used primarily: for data qathering--data read from an intermediate medium at user stations is transmitted to the computer or entered directly to the computer, elitainatinq the need for writing to an intermediate carrier. Data input operations at the user station are practically nonexistentt 12 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500060015-4 for issuinq referetrce informntion--the oomputer processes a request received from a user station and sends the response to the station. In th3s case the wluiae of infos~nation beinq retrieved usually exceeds the v~lume of information enteredt for solvinq message~switchinq problems--datrs is entered from asie user station and retrieved to another station virtually without processiags for c~m~uter control when the user station is employed as a computer operator console, and also for solvinq many other probiems. We will oonsider all the questions of remote processinq usinq the exa~le of the Ye5 remot~ data transinission system, where such systems as the YeS-7920, YeSTEL and others have already been developed. This system includes various devices and qeneral-purpose proqrams, standardized control a~.gorithms and pro- cedures, protocols that indicate the functional principles of different com- ponents of the system and also interfaces that define the conditions and patameters of inteqratinq renbte data processinq equfpment. The chief inethod of control in the YeS rea~ote data processinq systeta is the binary-synchs~onous method, which assinaes fulfillment of the follavinq ~ procedures: establishing and maintaininq synchronization in the data unit by means of synchronizinq symbols which may be included in any sequence of information symbols and also transmitted without them. These symbols are removed from messages received at the receiving station. Synchronization has three p~ases: the bit phase, which is intended to synchronize the siqnal conversion units, the initial symbol phase, in which at least twc synchronizinq siqnals are sent and which is done after establishinq bit synchx+onization but before the beqinninq of transmission of each block of data, and the phase of n~intaining symbol synchronization, where tw~ synchxnnizing symbols are entered in the in- formation flow each second during the data transmission process= - determininq the readinss of the data unit. After the synchro~?ization procedure, the traasmittinq and receiving statfons exchanqe synchronizing sym- bols in each direction. At the appearance of the first binary coaibination that differs from this syml~ol, the synchronization "time-out" is switched on. tTpon its oomptetfon, either the synchronization procedure is repeat~d or one of the three follvwing procedures is performeds set the datalink in a multi- point link, identify the station on oonununicationa channels to be switched or request receipt of data on segreqated channels--data transanission in the primary nr~de. The last procedure begins after the data 1 ink is established. Transmission is carried out primarily in blocks using resolvinq feedback= completion of data transmission. This procedure is perfora~ed when the signal for "end of transmission" is sent. Figure 12.1 shows a block diaqram of a simplified YeS remote data processing ~ system. The inteqration devices in it are data transmission multiplexers which have connectiona, on the one har~d, to the computer input-output channel FOR OFF'IC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2447/02/09: CIA-RDP82-00850R000500464415-4 rv~c vcr~~�~ha, a~~a: a~i~a.~ and, on the c~ther hand, to the user station through ca~munications equiPment. The data transmission multiplexer can be connected to the user station pera~an- ently by means of unswitched (seqreqated) co~nmunications Channels or tempora~ ily usinq switched channels. The data transmission multiplexer can be oonnected to one user station (s3ngYe- point coz~nection) or several tmultipoint cannectfon). The types of channels used and the types of connections have a significant influence on the processes of establishing and breakinq contact between data transmission multiplexers and user stations. Thus~ with multipoint oonnection on seqreqsted channels the multiplexer contact with the user station ts estab- - lished by sending special inquiry signals in th~ connnunications channel� When working on switch~d chann~ls, contact is established by dfa11in4 ttie nwaber of the desired user stations or CoQiputers dependi.nq on who is initiating the cc+nnection. ana . ~ KoKa~ ce~r~r . (3) - q ' AI/Q . ~ ~y~e~unn ~ NoMan 4 ~ . u~ Figure 12.1. Block Diagram of Sim131ified Remote Data Processing System 4. Multiplex channel 1. User station 2, Data transmission unit 5. Central psocessor 3~ Con~aunications channels 6. Data transmission mu2tiplexer The YeS remote data processinq system usea the five-element MT1C-2 code for operation on telegraph channels and tha seven-element ImI-7 code developed on the basis of the MTK-5 code for connections on telephone and wideband acmiavni- cations channels. Inforn?ation is translated from ROI-8 and DROI internal F'OR OFFICUL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 machine codes to transmission oodas partly by the softaare ai'id Partly bY the hardware o~ the data processinq multiple~oer. - The YeS remcte dsta pmcessinq system has tw~ types of devices: hardware and software. ~ All the remate data procesainq software interacts closely Mith the qeneral ooaQuter softMare and operates under its contml. ~ The software, its ~tructure and the soope of functions perfoxmed depand on the confiquration of the remcte processinq system and on the task it is per- forminq. Moreover, they are driented to particular models of oomputers and types of peripheral equipment� Remote data processing software is subdivided into applied proqraa? Paclcs that expand the capabilities of the operating systems and the basic software, which facilitates the work of rem~te processinq units and permits use~rs to write pro- qrams in symbolic proqramminq lanquaqe. The main con~ponents of the basic software are the basic teleooamunica~ions ac- cess methad (BTMD) and the general telec~ommim~.cations a~aess method (0'1'1~ID) ~ which is achfeved in the operatinq syste~n and disk operatinq system of the Yes ~ comguter and ensures the establishment of c~mmaunication with the user station, detection and oorrection of errors, oontrol of the buffer memory. aata conver- sion and so on. It is more difficult to achieve the basic telecon~aunications access method because it requires that the user, whea aritinq ps~oqremc~. have a detniled knowledqe of the tectnsical specifications and composition of data input-output equipment at the user station and that he also be thorouqhly fa- milfar with methods and modes of c7ata transmission. The user himself must in- clude the addresses of user stations and the inquiry sequeace in the proqrams , and perform many other additional fwi+ctions. Since the qeneral teleconaaunications acaess meth~d essentially precludes any effect of the remote data Prccessin4 harda?are sPecifications, this method psn- vides a higher level of software than the basic teleco~rmw'?ications access method. It performs all the functions of the basic methoc~ ai'~d also edits mes- saqe titles, aontr~is the organization of queues, establishes modes of inter~ action between the oomputer ancY the user station and so on. The qe+neral tele- voanaunications access method has been achieved in the operatinq system of the Unified Co~m~uter System and is used chiefly in hiqh-level reaaote data process- ing systems. 12.2. Remote Data Processinq Hardware Remote data processinq hardware includes da~a transwission equipment bY which the data processinq equiPmei'it is connected to the com~aunications aetw~ork, de- vices to inteqrate the co~nputer and the data trsnsmission equipment to ensure control of data exchanqe, matchinq of electtical siqnals and speeds and so on and user stations with various typee of peripheral equipment. Fiqure 12.2 shoars the classifiaation of hardware used in th~ YeS remote data Processinq ~ system. ~ FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~ h'UK ~l~'M'IC:IAI. U~~ U1~iLY The data traasmission equipment makes it possible to w~ork With both switched and segreqated teleqraph arid tele~hone chaanels as well as wirleband cc~ica- tions chanuels and physical lines. S~gnai.conversion devices, error-protection deyi.ces and nutomatic call devicus are includecl amonq the data transmission equipmei'it of the YeS system, as weil - as other data transLaission equipment. The m~st aidely used of these devices are signal canversion units. T'heY are in turn subaivided into the following: modems which convert digital binary signals from a computer or user sta- ~ tion to siqnals that cas be sent by telephone or wideband eommunications chan- nels and perferm the inverse conversion= siqna7. oonversion devices for teleqraph ccno~wzications used for operatinq on teleqraph lines to raise the level of the loqic siqnal and its outputt signal coaversion devices for connectinq lines to transmit data over dis- tances up to 10-14 kilaneters by connecting lines. The most widely ~d are low-level data conversion devices which convest diqital signals to low-level DC siqnals, which minimizes the�insignificant mutual effect of signals of different circuits. ~ � - . ~ a ~ 1 s~ 2 '�n ~n aer~ o Anq ' ynq5~ ~ao~6 ~ ~a~~~ a MnQ kn . rr yn,q ~�',,�~~p'"M"'y~�~"E' savn/t 10 11) 2 (13) (14) (15)` (16) , (17) Fiqure 12.2. Classification of Remote Data Processinq Hardware ~ ~Y: l. Remote data processing hardware 11. Signal conversion devices for 2. Data transm~ssion equipment telegraph channels 3. User station 12. Signal conversion devices for 4. Computeridata transmission conaectinq liaes ~ equip~nent inteqration devices 13. Automatic call devices for 5. Signal conversfon devices teleqraph lines 6. Ersor-protection devices 14. Automntic call devices ~or 7. Autoaiatic call devices telephone lines S. Data tranmaission multiple~oers Y5. Datn transmission multiplexers 9. Ccmnaautications processors 16. Proqrammable data transmission 10. Modems multiplexers 17. Remote data transmission multiplexer FOR OF'F[C[ ?L USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OFFICIAL USE ONLY The specifications of the various siqnal conversion devices developed for the YeS remote data trans~nission systems are shown in Table I2.1. Erroriprotection devices eacod~e and decode messaqes and detect errors in them. They use matrix or cyclic ~des which have the polynomials x16 + x12 + x' + 1 - or X24 + x23 + x~ + x5 + x + l. Error-p~otection devices with hiqh-level codes are used in cases which requfre very high data transmission accuracy and communicate in only a few directions. The presence of an error-protection unit co~nsiderably increases the cost of buildinq a remote da~a transmissfon system and makes constru~ction more oomplicated. In most cases, hawever, the ~ required transmission reliability is actsieved either by computer programs or by the haxdware and software of the user station. Table 12.2 presems the ~ specificatfons of es~ror protection devices related to the YeS remote ciata processinq system. The UZO-4800 is designed to w~rk with user stations not included in the nomen- clature of the Unified Computer System, while the UZO-48000 is for interconr puter data exchange and the UZO-2400 is for with AP-2 and AP-3 user stations. The a~utomatic call devices (AZO) automate the establishment of connections in " switched networks and are subdivided into units desiqned to work on teleqraph - (YeS-8063) and telephone (YeS-8061 and YeS-8062) caaimunfcations channels. ~ The follawing standards for unified interfaces have been established in the YeS resaote data transmission system to standardize the inteqration of specific units: sl junction for signal conversion devices to communications channels, S2 junction fcr signal c~nversion devices and automatic call devices to data processing equipment and 53 juaction for erroriprotection units to data pro- cessinq equipment. Devices to integrate computers and data transmission equipment consists of data transmission multiplexers and comQnunications processors. The data transmission multiplexers are the central units of the remote data processinq system and deteYmine its capabilities and configurations. These units provide: interaction between th~ computer and the user station through communications channels, with essential data conversion, partial buffering (the main computer meaory is used for complete bufferi~q), interference-sta- bilizinq codiaq and so onj execution of data control alqorithn~ that are achieved for each user ~ station and that contain procedures to establish and break the data link, for identification and inquiry and for data transmission. The order of fulfillinq the procedures is initiated in the data transmission multiplexer by a sequence of channel proqrams and instructicnso control of data transmission channels and equipment, i.e., loqic and electrical inteqration of the computer and the data transmission equipment are achieved. FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FOR OFFICIAL USE ONLY ,~a~a b ~ ~ a�~ a a a ~ ~ c " .~~i > " ~ > ~ ~ ~ - ~o m ;p a~i ,~~i a ~ ~ Pu A E ~ m = ~ = = - ~ - - m~ ~ a _ ~'~~xa '~�~a ~ ~ a a ~ c'[i m .~~i c _ ~ m _ = m - . ~ ~ m~ ~ ~ a ~ a, 0 0 00 O b O O .~T C~0 N N ~ ~ N N N ~t O ~ri , ~ 'o .-I .-1 ~ 0 p" ~ ~ ~ O O w w � U ~ d O O O ON N o O~ ~D +1 M pr N N ~O ~O rl N N u'f ~ - A E~ y ~ ~ ~ Ci ~O ~ �O ~ �.~1 W m O M C O t~+ O ~ d e f�+ e ~ s~+ ~ u p C-~'~ a~0 ~ ~ ~ ~ ~ v - - - v ~ - U ~ ~ ` y ~ ~ ~ ~ ~ ~ I 1 N ~ ~ ~ Ol 01 ~ d m c~0 wx ~ ~41 ~41 ~ !C i6 ~ a i6 ~C r1 fl iC !6 .-1 ~ ~ 41 QI CL 'd ~JI N O. 41 N N C~ ~ ~ a a~v ~ a a~o a a ab m ~ A A v~i fi ~ A a ~ b a w U ap C ~ V J I~i M ~0 ~ ~ 10 M r~l rOl ~ . CI a~ a+ q~.1 ~ 1 M C i~ G ~ ~n ~ ~ ~ awe d ~ .�e ~ ,.eoi ~ a ~ ~ a ~ W O ~�1 00 C7 00 C~ 9 1+ ~.1 ~ ~ L 1~1 Y O 7 U 41 .-1 G C1 O e0 41 ~0 d Gi ~0 al W � C M� ~ a1 7 N.~I 00 ~r~l 00 rl ~r~l 00' . ` N ~ y a) ~:i: U~ ~ fA .'L' V~! V~] f~ y ~ ~ ~ r-~i fr O O O ~ ~ ~ .~-1 N O~ ~ ~ .~G o~0 0~0 0~0 C~0 e~D C~0 a~0 0~0 O .j ~ H ~ W Vf Vl V~ t!~ t%] V! N ~j ~ ~ 7+ 7+ i~ ~ ~ 9~+ ~ y U 0! ~ O O O O r�~ E O ~ ~ � g ~ N N ~ ~ ~ a ~ ~ > ~ m ~ > ~ ~ b b b ~ ~ ~ ~V D N ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 18 FOit OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500060015-4 i ~ . . m ~ ~ a~ a . oy,~ a~, ~ ~ ~ m H ~ ~ H , O ~ O = N w ~ ~ . n ~ ~ ~ a ~ = - ~ ~ . ~ 0 ~ ~ a e ~ ~ > w ~ m ~ ~ ~ c ~ a~o v ~ O U " Q N O O N ~ ~ ~ N N 41 rl H . ~ ~ ~ ~ ~ ~C~~. a$o ~ m ~ N ~ H N > 19 ~ FOR OFF'ICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 h'UI~ JFN'K:IAL U~r. UNLY The YeS remote data transmission systea~ assumes several types of data trans- ' - mission multiplexers that differ by the nianber and variety of units of data traasmission equipment, user stations and communications channels. The com- puter in all data transmission multiplexers is connected to the multiplex : chanr~el and the structure is similar, consistinq of a computer inputroutput : channel inteqration device which perfaxms parallel data exchange between the ~ data transmission multiplexer and the computer, a tw~o-channel switch that per- mits operatian with tw~o multiplex channels and different m~dels of YeS comp~ i ters and line adapters that take into account the special characteristics of user stations that are being connected. The ntanber and composition of liae adapters depend an the type of data transmission multiplexer and the configur- ' ation of the remote data processing system. Table 12.3 presents the specifications of data transmission multiplexers that ~ have been achieved with the Ye5 rem4te data transmission system. Al1 the mul- i tiplexers are cannected to the multiplex channel of the computer throuqh a , standard input-output interfacej they are oonnected to the data transmission equipment through an S2 junction and directly to telephone channels through an Sl junction. . The MPD-lA (Yes-8400) data transmission multiplexer permits up to 15 different ~ user stations to be connected to the coaiputer through cammunications channels. ~ These stations can be the AP-61, AP-63, AP-70, RTA and any model of the YeS~ computer which includes an MPD-lA. The transmission speed over physica.l lines and telegraph coamn~nications chaiuiels is 75 bauds= it is 200-4,800 bauds over telephone channels. Telegraph communicatiuns channels and physical lines are connected to the mul- tiplexer either separately or through a signal convsrsion unit; telephone channels are connected by a Modem-200 or l~bdem-24~~� 1�?11 inoomin4 and outgo- inq connections are automatic. The only exception is establishing outgoinq connections on switched teleqraph lines. This operation is performed at the in tiative of the ca~puter by the operator, who manually dials the awaber y3- trieved by the computer. . The following line adapters have been developed for the MPD-lA: AD1 for work- ing with page-printinq telegraphs throuqh switched telegraph comaaunications channels, the AD-2 for workinq with the AP-70 through segregated teleqraph channels, the AD-3 for working with the AP-61 and AP-63 through the Modem-2400 and seqreqated telephone channels and the AD-4 for workinq with the AP-1 and AP-70 through telephone channels and the Modem-200. These adapters pr~vide a - semiduplex data exchange mode. The AD-5 adapter coamnunicates betw+een models of YeS computers in the duplex mode. Zhe AD-6 synchronous adapter is used for workinq with all synchronous user stations on unswitched teleqraph lines. The MPD-lA has an adapter connection unit (BPA) for connecting the line adap- ters that commaunicate�with the channel inteqration unit and an adapter syn- chronization unit (BSkhA) for control of the adapters. Fiqur~ 12.3 shows one possible confiquration when using the MPD-lA. 20 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~ F'OR OFF[CIAL USE ONLY o g 0 W O 41 ~-1 N - ~O ~ ~ t 8 .G rl .C 8 ~ O .~C C~ O~ O~ O~ O~aO ~U L ~i~ E~~t ~ ~ ~ ~ya a a ~o ~ � u ~ ~ w b~? p, �i a a~ a�'i ae0i m..mi ~ ~ ~ E E ~H ~ ~ 1 ~ ~�~j o a p m ~ ~ ~-~i ~ ~ ~ a a a . i~ A A ~ m ~ 1 i 1 V ~ d 7+RH m o~s a�+ x a�~ a� x ,e ~ o~~ a s~ ~ ~ ~ ~ , ~ 3 ie ~ ~ w~~ a~~ 3 3 ~1 00 + ~ ~ ~ ~ o ~ t~i~r ~t~ ~ ~ oeie v~ ~~~a t'i'a�~a~+ a - ~ . ~ 'd ~ Iil~ F'1 u Id M ~ O V ~ .~tl~ r~4 . ~ O ~i . W O ~ p ~ W'O (0 N Cy ~ w w Q m G.a~O ~ ~N.7 ~O~ap 00 a ~j ,Q w w w w Q w w ~ V 10 O Orlr~l �NN~T ~~T E~ai ~ N u~1 N W V ~ ~ ~ ~ a~ ~n u ~3 ~ ~ s+ N C: 00 y ~ f~A i~A ~ ~ ral D r'7 d1 ~ N u'1 O . C-a~ O O D y �~j~ a'~'o a'~~0 U Q i~ ~ O O O O O � a e�~ .~t a�~ a�o ~ N N N N � a a ~ a ~ ~21 FOR OF'F[CIAL U3E ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 r'UK UNI~'1(:lAL w~ uaLx , a~ ~k- m A1 p ~o = G ~ m ~ U . w ~ ~ ~ A U m ~ 3i a~-' G~ 1~i ~~~-1 ~ V ~ ~o ~ c ee w u . a~i ~j ~ .x . C~ ~ ~ " ~C ~ ~ a _ ~ L ~ a _ ~ ~ ~ ~ c�o a�o o�o a $ w w w w w N ~ m't7 ~7 Q ~ N rl N d H~ ~ ~ ~ � N a �w . N1 ~ N ~ ~ .w fD ~ f0 �w m ~3 ~ 41 rl ~t CI ~ Cl ~O ~ ~ x o ~.e i i.c .e i ~ a ~ y~j �w ~ ~ rl ~ ~ � 0 ~ ~ ~ l0 N ~O �w rl ~i7 00 a ~ ~ y ~ ~ 6 d O. e~ ~ ,O ~ ~o ~1 ~-~I ~ a~i ~ ~ ~ ~ l~ ~1 ~i ~ ~ w ~ ~ L . pp �w C CI C 0~0 �w N.~7 O ~ ~ p1., 0) N Ia~ 'O ~ N ~ ~'r1 ~ v1 ~O ~ � ~ WC 1 6 a, ~ C~ a~+ 1~~ 1 1~ w a ~ a ~ ~ �w ~ ~ w ~ w w ~ �w 5~.1 �w 1~.1 W ~ rl ~D ~ .-1 e^1 ~O N r-1 ~ 41 ~ 41 rl ~ r-I QI p ~ ~~d ~~~d s~ ~ o H m e ~ a o~a?~ ~ c~e o W N C) ~~�1 u1 ~ ~ N r-1 N O ~H ,ar,, y 0~0 ~ z c�~ a~"? ~ m o ~ ~ y ~ . ro U ~ w ~ ~ ~i ' V ~ ~ ~ - N t~l ~ ~ ~ ~ . ~ ~ ~ ~ ~ U a a ~ ~ ~ ~ ~ N N ~ ~ ~ ~ ~ ~ ~ ~ ~ C? p O O O O N ~ O a~0 0~0 . 0~0 0~0 ~ ~ - U I 1 ~ ~ N N d y y d W 41 ~ ~ ~ ~ ~ ~ .~1 'ti b d b d b " ~ ~ ~ ~ ~ ~ 22 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500060015-4 The MPD-3 (YeS-8403) data transmission multiplexer provides co~nmtimications ~ between the YeS-1020, YeS-1030, YeS-1050 and YeS-1060 oomputere writh four user - stations, as which the AP-l, AP-3, AP-11, AP-70 aad the page-printinq teleqraph or the YeS computer with the MPD-3 can be used. r-----------1 ~ (10 ~ i . - . ~ . q,q.~ ( ' yIJC . J xoNOn c Asu PTA (14 ~ . I A,Q-2 � yAC ~aeA~~ - R/!- 70 3en 6cx ~ .I _J KaNOn ari (15) ~ 2, ~ ~'�J C iraNOa~ au ~-~s (16) ; . ~5 6~~ A~-~ ~ .r~?-~ ma~ qn-io ~ ~ . ~ ~9) ' 'r~ (13) � ~Mnq_~A (I~D-iA) ' Figure 12.3. Configuration af YeS Remote Data P~roc!ies~.�iq S~stem vsinq MPo-lA ~ IGey : _ 1. Computer 2. Input-output channel-computer integration unit 3. Channel inteqration unit 4. Adapter connection unit 5. Adapter synchronization unit 6. Line adapter , ~ 7. Signal conversion device 8. Modem-2400 9. Modem-200 10. S~ritched telegraph communications channel 11. Segreqated teleqraph communications channel . ' ' 12. Segreqated telephone connaunications chanael 13. Switched telephone conmunications channel 14. Paqe-printinq teleqraph � 15. User station . 16. Or . The described multiplexer has the followinq types of line adapters: TA-1 start-~top adapter by which the AP-1 and AP-70 user stations are - connected to the MPD-3 over switched and segreqated telephone channels throuqh the Modem-200 and AV[~TF autvmatic call device= TA-2 start-stop adapter that f~cilitates cammunications of the MPD-3 over segregated telegraph channels with user stations in which telegraph equipment has been installed. Only hardware control of the infarmation being transmit- ted is exercised; SA-1 synchronous adapter'through which the AP-11 or similar MPD-3 are connected to the multiplexer over seqregated telephone channels. . 23 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~ Cou~aunications is es~ablished through the Modsn-2400. The error protection method is a cyclic code with qeneratrix of polynomial x16 + x12 + xs + l~ AA-1 asynchronous adapter through which the UZO-1200 error protection ' device and the Modem-1200 connect AP-2 and AP-3 user stations to a qive~c mul- tiplexer over switched and segreqated telephone oone~nunications channels. The reliability of the transmi~ted information is increased by usir?g the error protection device. The MPD-3 can be supplied with various types of line adapters to provide a semiduplex data exchanqe mode. The following line adapters--DA-1 when v~rorking with a user station not included in the Unified Computer Systeai and DA-2 that supports intercesnputer exchange--axe used to operate in the duplex m~de. In- formation is trarismitted over coamunications channels at speeds of 50-4,800 bauds. The oonfiguration of the remote data transmission system usinq the - i~D-3 is shawn in Figure 12.4. ~ ' ' ~ (13) � ~ II~NM uq~r~7 u l~de- an-~ - ~ TR-f A~ ~rQ- R-7D t18 ) ~ B~ � . ' 6cx rA-2 ~ ~ ynF- L~"den,ro v,+~ su~~~+�w vra (19 ) - aen ~__J . ~l ~~~2) a e 1 e~~aa,a~wi~ rsu 15 " An-r~ ; 6 ~j C'~-f ~ _ xaMa~ c saw~l3) ~ ' (8 A~-t ' y3a-uoo �aw~,a n,re- "n-a ~ ~-y~-? --------J ~12) (16) ~ a~A~tr�" . = Figure 12.4. Configw:ation of YeS Remote Data Processinq System Using MPD-3 ' Key: l. Computer 12. Error protection device 2. Channel integration unit 13. Switched and segregated tele- 3. Channel connection unit phone co~cnunications channel 4. Adapter connection unit 14. Segreqated telegraph coaimuni- 5. Adapter synchronization unit cations channel 6. Telegraph 15. Segreqateai telephone coamuni- 7. Synchrcnous adapter cations channel 8. Asynchronous adapter 16. Modem-1200 9. Modem-200 17. User station l0. Signal conversion unit 18. Or il. Modem-2400 19. Paqe-printing telegraph The MPD-1 (Yes-8404) data transmission multiplexer provides oonnection of user. stations ovar 32 (with possible increase to 64) swi~ched and segregated tele- - phone channels and unswitched telegraph chann~ls, as well as on physical lines. The speed of transmission can be 50, 75, 100, 200, 600, 1,200 and 2,400 bauds. - 24 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FI The MP~-1 has three types of adapters--the TA-1 works with the AP-1 and AP-70 user statfons through switched and unswitched telephone channels, the TA-2 _ w~orks with telegraph equipment and the TA-3 operates with AP-61 and AP-63 user stntions over unswitched telephone communications channels. The MPD-1 includes a memory unit containing two modules with 32 addressable cells each, which contain the line control words. One module has a capacity of 2,112 bits and is connected to TA-1 aa�3 TA-2 adapters and the other, havinq a capacity of 2,240 bits, is connected to the TA-3 adapter. The MPD-2 (YeS-8402) progrannaable dnta transmission multiplexer is designed to orqanize large systems within the Unified Computer System.. It exchanges data between computers and all typea of data transmission multiplexers and user stations included in the YeS remote data processing hardware. The MPD-2 permits connection of 8 to 176 communications lines with spacing of increasinq the number of coma?unications channels equal to eiqht. Unswitched and seqreqeted telephone ca~araunications channels and physical lines can be used in this case. The transmission speeds are 50, 100, 200, 600, 1,200, 2,400 and 4,800 bauds. Unlike the considered hardware data transmission multiplexers in which the nomenclature of the user stations is always limited, the 1~D-2 permits the connection of different types of user stations. This advantage greatly in- creases the flexibility of the systeza, contributes to expansion when necessary and increases the number of connected user stations. New user stations are oonnected by rewritinq a proqram into the I~D memary without makinq changes in its circuitry. T'he MPD-2 includes a disk memory unit with capacity of 4,096 72-digit w~ords and a microprogrmriming control unit. - Using these devices, the multiplexer performs fuactions related to execution of specific exchanqe alqorithms. Besides these devices, the MPD-2 oontains an internal storage device where cantrol and inform3tion line words are stored for all con~nunicatfons channels. The YeS-8421 ~maote data transmission multiplexer concentrates 20 unswitched telegraph coannunications channels operating et a speed of 50 bauds and havinq user stations at the ends into a single unswitched telephone line having a speed of 1,200 bauds. The use of this multiplexer permits the number of com- munications channels to be considerably reduced and a significant increase of their utilization efficiency. One of the possible configurations of a remote data processinq system using the remote data transmission multiplexer is shawn in Fiqure 12.5. , Since the functions of the remote data processing system related to analysis of headings, establishment of queues, messaqe editing and interrogation of user station status are performed by a computer, further developaient of the - YeS remote data processing system assuznes the development and introduction of special connnunications processors and the related redis'cribution of remote data processing functions. 25 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 r~K t~rr~a.~w~ u~c ~riL� ; - ~ BMaQ/1~NNNli ; ~ ~eae~N~n+ An-t ) . -T00" o~ q/i 7D (10) , ~ ~1I~G~ v Be+aena+aw~ ~eae- ~p � nawas N~ q~ ~ ~ 3BM Hll~q"? yM~A M�~'~ ~onnu KoNO -?00 RM64 2 ' 3 4 ~S~c ~ou (4 3 . ~a~ ~neo wN . PTA !If?MO dlt Fiqure 12.5. Conliquxation of YeS Remote Data Processinq SYstem Osin9 Remote Data Transmission Multiplexer ~Y ~ 1. Computer ~ 2. I~D-2 3. Remote data transmission multiple~oer 4. Modesn-1200 5. Segreqated telephone oommunications channel 6. Modem-200 ' 7. Segreqated telephone oontau~.cations channel 8. Seqregated teleqraph communications channel 9. User station lo. or 11. Paqe-printinq teleqraph . ~ The use of coamunications processors instead of data transmissiw'i multiplexers contributes to a redu~ction of the c~nputer load, an increase of the flexibil- ity and reliability of the system and also to the capability of usinq them as a message-switchinq center in oomputer networks. In nest cases c~uuaanunicatians pmcessors are minicctnputers supplemented with special software. Line adapters acoomplish inteqration functioas of the minicomputer to the central crnaPuter and transmission devices. COPYRIGHT: Izdatel'stvo "Statistika", 1980 6521 CSO: 8144/0540-A 26 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 F~OR OFFICUIL USE ONLY GENER~L DiITA ON OO1~UTffit CONFIGOR7ITION lYbscaw 8V!! I VYCHISLITEL'NYYS SETI in Russian 1980 (sigtrsd to press 20 May 80) pp 61-67, 117-118, 122-124, 131-134, ~.45-148, 157-158, 224-250 [~ocerpts from the book "Ccmputers and Dom~uter Netuorks� by Vasiliy Nilwlayavich Kriushin, Inna Nikolayevna Buravtseva, Nina Mikhaylovna Pushkiaa and l~ina Grigor'- yevna Chernyak, Izdatel'stvo "Statiatika", 18,000 copies, 328 paqes] [Excerpts] Chapter 4. 4. 3. G~omposition atid IYlain Chara~cteristics of Oaf fied (7o~eputer System as~ iKodular ~onQtt~er Equipment Systams The uaified con~puter systeta (YeS EVM) is a family of proqram-ooonpatible models of third-qeaeration eomputera desiqaed to solve a vide raaqe of scieatific and tech- nical, ecaa~amic, infoxaation-loqic and c~oatr~l problems. These machines are or- iented toward their use ia automated cc~trol systems. The LTnified Computer System w~s developed Mithin the fra~ne~ark of vooperation of socialist member countriss of CSMA. The first uait (Rya~d-1) of the Zfiified c7aa~- puter Systeai iacludes sev~eral models produced in larqe seriss (Table 4.1). Ia- dustrial production of the first machines of the Onifisd System i+as bequn in 1972. Tabla 4.1. - Models of tlnified Com~uter System Dev~slopfaq Country YeS-1010 Hunqarian Peoples Republic YeS-1020 USSR aad Peoples Republic of Bulqaria ~ YeS-1021 CSSR YeS-1022 OSSR YeS-1030 USSR and Polish Peoples Republic YeS-1032 Polish Peoples Republic YeS-1033 IISSR YeS-1040 ~ Germaa Democratic Rapublic YeS-1050 ~ OSSR The computer hardware of the ~aified System includes processors with different speed, selector and multiplex channels, an aggreqate of standard interfaces 27 F~OR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OFFICIAL USE ONLY between channels aad the contrc~l devices of external devices and a larqe nomencla- tiae of peripheral d~evices of different 8esiqnation (see Figure 4.2). The mafa featnres of the camputers of the Unified system ares proqrmn compatability "from bottrna to top" of individual computer models that easures the succession of proqrams when switchinq froaa one modal to anothert standard iateqration of external devices Mith input-output chanaela that per- mits a large avabar of different external devices to be ca~nectad= capnbility of combining several ccmaputers iato a sinqle aystemf ~ capability of aperatinq in contral systems in real time Standardization aad unification provided by a system of standards became necessary when developiuq the Lfiified Computer System. The system of atanc3arc]s or system of normative and technical documents of the Unified Co~puter System is underst.ood as the complex of interrelated standards that establish the aqgreqate of the norms, regulatic~s aad requirements, fulfillmeat of ~vhich datermines the modular design of ~omputer models of the Unified S~istem. For example, state standards on computer equipment (GOST 16325-76 "Geaeral-Purpoae Camputers. General Specifications") es- tablish six clasaes fa productivity for a family of camputers, deteraiininq for each class the lawest capacity of the internal storaqe devices and the functfons per- formed by the coapuCe~ The models o� the Unified Coaiputer System can be divided� into classes as shawn in Fiqure 4.3. Tfie main characteristics of the computers of the tfiified S~?stem (the Ryad-1) are presented in Table 4.2. Each of the models of the Unified Computer System has (c~nQulsory) a mii?imum standard oomplement of hardware ac7equate for functioning of the correspondinq op- eratinq system. The mfnimnm com~osition of the equipment of the first unit of co~nputers of the ifiified System is' preseated ~n Table 4.3. The priaciples on which the family of the Onified Computer System was dev~eloped permit the capability of each of the oomputers to be expanded by increasinq the internal storaqe capacity, increasinq the number aad chaaqinq the nomenclature of external devices and ao~u~ection of a seaond ps+ooessor. Proqraa~ and informntion compatibility are fully retained on expausion. Auxiliary fu:?ctional devices cc~?tribute to an i.ncrease of system productivity, ori- entatioa toward specific areas of application and to providiaq increaaed operatinq requirements. The oombination o~ hardware arid software of the Onified Computer System easures op- eration in all m~3es usually required by users: batch provessinq, multipr~qram, real-time, dfaloque a:?d time-sharinq. The m~dels of the Ryad-2 tAiified Oomputer S~?stem inclvde the followinq: YeS-1015 (Hunqarian Peoplea Rspublic), YeS-1025 (CSSR), YeS-1035 (CSSR) and Paoples Republic - 28 . FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~n ~ o o`~'+ + + + + + + + ~ol o r�~ ~ K ~ 1~ P O O ~ ~ .~-1 ~ N O ? ? ? ? ? ? ? 14 ~ ~ ~ ~ O ~ ~j 7 'al 000 .N-1 + + ? + + + + rl r1 f~ Q If1 ~ U p . ~ w rr ~ O p ~-~1 O N U1 O ~ .-1 ~ erf ~ P1 N N e} u1 � 1N ~.-1 .~-1 .-1 R~ N N ? 1 1 ? ? 1 + 0 w ~ N V Yi ~ ~ i~~l O O ~ NI p ~ 1~1 .~V~1 ? 1 t ? ? 1 + b ~ p Oa i~ ~'j ~ 'a~ o ~ ~ ~ i i i + + ~ + m ~ ~ ^ ~ ~ � s~ a a a ~ _ ~ ~ ~ ~ - ~ ~ ~A ~ ~ a~ M ~ ~ O S ~ +i . ~ 1~0 p ~ ~ J O O M ~1 pV i~ ,r~ ~ ~ i~ yo~ ~ ~ Wa o ep~ 4 b' s~'? p~ oD, ~o tj ~FI ~ ~ ~ O ~ a ~ a~ ~ o? ~ o�a o~o o~�a~ o0 ~e ~ ~ ~ ~ ~ � ~ ~ ~ ~do, ~.~a~ + ~ w q ~i o o Ga o ~ ~e ~ 4 ~ ~ o ~ ~ ~ ~ ~Q~ ~~q ~ ~ W ~ H , r7A U a ~ ~ 29 ' F~OR OFFICIAI. USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500060015-4 tVK ulr~a-~wa. a,+~a vi~~~ ~/l O O .~1 ~1 ~0 O N O O ~ n ~ ~ O ~A ~ ~ o\ ~n ~ ,C . O ~ ?1 ~ ~ C .~1 _ ti N O N ~f1~D ON 00 O 'd O m ~ N~ y~j a a m ~ ~ � n ~ ~ ~ b a ~ O p ~ Nt~A O~ m.~1 O W 1 u1 ~ � ~ ~ \ . ~-i ~-1 ~01 . ~ ~ ~ ~ ~ ~ ~ b N m ~ W ~ O O O ~ ~ N~ 01~~1 O~ O ~-1 O ~ ~ � ~ 1+1 W ~O . ~ N n ~ ~ .a ~ ~ ~ O ~ a N ~ ~1 tp O N O.-1 O ' .~~1 N N ~ � N~ sa � ( ~ ^ ~ ~ ~ . N O ~ m V M ~ .~~1 \ ~ ~ ~ i~ O ~ ~ ~ .q.~ U ~ ~ ~ p M .~-1 i ~ b ~ ~d a b~ a~ ~ ~ ~ ~ o ~ ~ ~ m o ~ - +~i m ~ ~ ee~D'jy�� am~ o\ y~ ~ v . ~ b ~ ~ ~ ~ ~ m . ~ ~ %~8~~ ~ ~ o~ b . v ~ o . t~ V ~ ~ 7 ~ O m ~ O m ~ .~a1 SJ ~ ~ U 'd w b ~ ~ ~am ~r~ ~ ~ o.~~ ~ ~ ~~~j w ~ ~ s a~o u~i q 30 FOR Ol~'F[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 F~OR OFF7CIAL USE ONLY ~ . ~ ~ ~ ~ ~-1 ~ ~ .-1 rl rl rl rl .-1 rl !~1 N 1 N ~ D O Z ~ ~ ~,q n, 1p .~1 P N n1 O N l~ O O frf O ~A rl U1 ~-1 rl N rl N!~ rl rl y ~ N O O 1f1~f1 00 00 O 00 1 H t+f ~f1 1~1 ~e1 1f1 ~O ~ e~ P e~ O~ O+ 1 ~ ~ 1 1 1 1 1 1 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ r r O � � � r � t O ~O i~ ~ ~ ~ ~ i~ ~ i~ ~ ~ ~ � ~ ~y nI V~ rl ~-1 rl rl rl rl rl .-1 1 a a o~ I N N t0 1~ ~ l~ N N M!~1 1~ O O .-1 fA N N 1 IA ~-1 1A rl ~-I ~V ri N t~ rl N ~ e ~p N o 11 O O 1A tf1 O O O O O O O 1 "O ~ N erf i1 ~ff 1ff U1 U1 10 ~O P P f~ O? O~ 1 p 1 1 1 1 1 1 1 1 1 1 1 1 ~ V y y a p s fA tA N fA tA M tA fA fA ~(A ~ ~ ~ V ~ ~ i~~ ~ i1 y ~~p ~-1 N i~ N 1A m rl ~-1 N rl rl ri ~V rl rl rl ~ O F~ O 'lar S~ 1 O U1 N ~A O O? rl ri ~''f N M~`I ~ O O~ ~ o N ~[1 O .~1 ef ~A rl ~f1 .~1 r-1 eV .-1 n ~ ~-1 N 1~ ~ ~ ~ N ~ ~ t~f1 1~A 1~A 1~A 1~0 ~ ~ ~ ^ ~ O~? ~ ~ UI v1 ~ tif tA ~ N v1 O O O O ~ U ~ ~ ~ ~ ~ ~ i~~ ~i~ ~ i~ - 'C! ~1 W O~ .-1 ~-1 .-1 Nf N ef .~i ~-1 .-1 .-1 .-1 rl .-1 N N 1 +i 1'~1 ~ ~ ~ w N p t+1 O O ~ O? .-1 ~-1 N N O eV O O O 1~f O 1+f Nf ~A rl 1f1 rl .p-1 N ~-I eV rl N ~ ~ N ~~+f Q ~ ~~f1 N` ~ ~~f1 ~O 1~0 ~ n n O~? ~ 1 ~ N ~ f!~ N N tl~ ~ ~N ~ ~ ~N ~~fl ~ a1 N ~ r ~I O ~1 O � O O~ � O O � O 'l7 ' ~ .~1 D+ ~ ~ ~ ii ~ r.l O rl rl N d' rl rl rl rl rl rl rl N N 1 ~ N ~ ~ O Z ~ i O O ~C O r-1 rl eV N O tV O O O ~ fA N N I 1A ~-1 1A rl rl ni ~ N ~ ~ N +1 O dl O N ~ O ~ I~A 1111~A ~1~0 ~ _ a ~�~I ~ ~ am ~ ~ ~ ~ ~ ~ ~a, ~o a�~ ~ a� a�~ ~ a~ ~ a~ a� ~ a� a� ~ U +UI fl 1 m " ~ ~ a ~ W a ~ ~ ~1 ~ A ~ ~ O ~ t~ a A N . ~ x ~ ~ o ~ bx �a �.~'a ~ ~ ~ ~oc~~ ~ �c~ e~ o,~ ~ a~ gn a~ g+ o o ~ a�o ~~.~i ~ m m o~ a~ o o b � b ~ ~'~d o ~'c~ ~m � ~ ~ �i~~ aa~ a~a~ ~~S ~ ~ a ~ ~eo ~a $ ~ ~ ~~i ~a~ 31 . F01t OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FOR OF'FICUIL US~ ONLY of Bulqaria), YeS-1045 (OSSR and Polish Peoples Republic), YeS-1055 (G~CR), YeS- ~ 1060 (OSSR) aad YeS-1065 (OSSR) [17j. . Th~ main specifications of m~dels o~f the Ryad-2 Onified Oomputer S~stem are pre- aented in Table 4.4. Table 4.4. 1~lain Characteristics of Ryad-2 lrbdels of the Qfiffed Oomputer System Developed is the OSSR Pr~ductivity Storaqe Oode Producinq Oountzi? thousaad aPs/s ~~~itY.~Yt~ YeS-1035 ~ OSSR, Peoples Republic 140- 160 256- 1,024 . of Bulqaria YeS-1045 USSR, Polish Peoples 540- 880 1.024- 4,096 Republic YeS-1060 OSSR 1,300-1,600 1,024- 8,192 YeS-1065 OSSR 4,000-5,000 2,048-16,324 These mc?d~els retain the pmqram coiapatibility with Ryad-1 oomputera of the Dnified System. Tl~e main differance of ~yad-2 models iacludes.a naw ~ent bnse--larqe inteqrated circufts, aa increase of speed, a:? increase of storaqe capacity, qreat- er reliability aad signifiraat expansion of functional ca~pabilities (dynamic mem- ory distributia~, ocn~lexinq of models aad multiprocessor systems and so oi?). The modular ooaiputer equipment syst~m (ASVT) Was developed to solv~e problea~s of preliminary processinq of eo~wmic iaformatfon, calcnlatfaa~ of anqiaeerinq and evonomic iadicators, processinq iafosmation in information-measuriug syetems and ceatralized monitorinq systems and contr~l of produc~iaa facilities. The ASVT is a set of modular~ devices with unified external cananmicat3c~s, from ~/hich different computer systems aith qiven tecl~nical parameters can be ooa~fiqured, beqiani.nq with the sinQlest infosinatian qatherinq systems and endinq aith oo~mplex multiprocessor c]ata pracessinq systems. ~e complete set of com~emications-standardized uaits and devices permits one to take into accouat the m~st diverse requirements of ca~trol systems and to develop complexes that have satfsfactory redundaucg? from the user's viewpoint. The char- acteristic feature of ASVT is that their specifications can be varied over a wide ranqe when developinq different systems. Moreover, the m~dular system permits qradual modernization by replacinq iadividual 8evices by more ~dern devices. Equipment redundancy due to stanc3ardization of functional ccmm~micatioas uaits is repaid by serial production of a coaq~uter equipmeat system. ASVT can be divided into aix qxoups acwrdinq to functional designation of the device: centra~l contr~l and information processinq devices--proceasors= informatio~ storaqe devices--internal aad external storaqe dev3cas= � 32 FOR OFFICUIL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OFFI~'! USE ONLY ~ facility-com~aunications ~svices, the canplete nomeaclature of Nhich ansures ~ info:mation communications of the canputer systa~ desiqned on tha basis ot ASVT with any of tha sensors and serwmschanisms serially produced in ths OSSRs personnel oo~auaicati,oas davices of ttse operational praiuctia? contmi systam, intormation carrier iriput devices and output devicxs: extrASystemg co~mn~aications lina output devices (teleqraph, telephone, radio relay and so on)= intrasystems caaawaications dsvices that pravide infos~sticai exchange betwsen modular devices Nithin the system. The secoz~d vuit of the ASVT (A.SV7~M) Was desiqned on a microelectz+ouics base (in- teqrated circuits) and permits dev~elopment of coa~plexes of nachines and devices of different classes and also has the capability of c~omm~uiicating the Onified Computer System. The ASVT-M cantains M40 and M6010 automatic microproqram devices, M400, M5000, M5010, M6000, M7000 and 1l4030 diqital computers. a larqe nuaiber of external syste~ms devices, facility-conmunications devices and diqital and qraphic iaformntion dis- play devices. Models M40, M400, M6000, M7000 and M6010 have hiqhly effective devices for camnun- icatinq wfth ootrtrol facilities and are uaually employed ia producticu Ps+oaess oontrol systems, sci~ntific experiment automatioa 8ystems and so cn. Thase ma- chines operate primarily in real time. Models 1K4030, M5000 aad M5010 are characterized bY hi9her Productivity compared to earlier produced models, an exteasive instruction system and the preseace of al- phanumeric processing devices. They are also desiqned to operate ia real time and in this case coaomuaications with se~sors are m~st frequeatly carried out throuqh low-level systems desiqned on the basis of M40, M6000, lt7000 and 1K6010 computer complexes. ASV'1~-M devices permit multimachine oomplexes to be constructecl fsoa~ camputers of different classes, for example, I+i4030 oo~nputers of the SM $VM I~~~~~~ ~11 Computer Systeml. The most Productive cos~lex of the ASVT-M is the M4030 oomQuter ccm~lex. The M4030 cannlex is called upon to camprise the tachnical basa of ASU fn combination with devices and automatic equipmeat of other m~dular oomple~ces of the state instrwnent syst~n (GSP). The I~14030 psevides the capability of realiziaq multimachine hierarchical contmi sy~tems based on modsle dif~erent in producti~?- ity and desigr?ation and contained ia the ASVT-M. The M4030 model is designed for nse as a cantral mnchine: in the ASL1P [Autaaated prcaduction control system] to solve a ccmplex of prob- lems (control of production, optimtan orqanization of it, enqfaeerinq and ecanomic planninq, bookkeepinq and acoountinq and so on)t 33 FOR OFE7CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 rv~c vrr~a.uw a.o.: vivi.~ ~ ~ production process ASD (toqether with tha International Smal' Computer System)t in scientific and physical experiment automntion systems and so on. The confiquration, equipment and software system of the M4030 nadel ens~?~re its com- patibility vrith the Unified Computer System in instructians, data fora~at, internal codes, interface and user programs i.a assembler and hiqh-level languaqes havinQ translators in the DOS (disk operatinq systeaa) of the Unified Oomputer System and the DOS of the ASVT. The processor of the M4030 model has the universal instruction sYsten of the ASVT and the Dnified Cctnputer System expanded by introducinq awciliary transmission instructions, camparisoa of larqe information files and shiftiaq and roundinq off decimal operands. The average praductivity of the processor (accordinq to Gibson) is 100,000 operations per second. The internal storaqe has capacity of 128-512 Imytes and cycle time is 2 . microsevonds. The inputwutput devices and external stores are connected to the processor by three selector and or~e multiplex channels. The multiplex channel transmits data in the monopole an~de at a speed of 140,000 bytes per secoad and in the multiplex aade at a speed of 50,000 bytes per second. The selector channal has maximwa~data transmission speed of one million bytes per seoond. The M4030 contains a standard aet of peripheral devices: mnqnetic tape and maqnet- ic disk external storaqe, punch card and papertape input devices, printout, punch card and papertape devices aad displays. - The processor of the M6000 model is the minimwa baseline coaifiguration of the com� puter oomplex configured from modules (57 types) atid 15 differeat complexes can be developed on its basis depending on the sphere of application [Z]� domputer Storaqe Devices Table 6.1. Internal Storaqe Devices (OZU) Storaqe Cycle Capacity, Diqit uter Device Code Time, us ImYtes CaPacity, bYtes YeS-1020 YeS-3220 2 64- 256 2 YeS-1022 YeS-3222 2 128- 516 4 YeS-1030 YeS-3203 1.25 128- 256 4 YeS~-1033 YeS-3207 1.2 256- 512 4 YeS-1040 YeS-3204 1.35 256-1,024 8 YeS-1050 YeS-3205 1.25 256-1,024 8 BESM-6 2 64- 128** 50* ASVT-M A-211-8 2.5 4 2 M4030 2 128- 512 2 ? Bits for the BESM-6. wnrds. 34 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500060015-4 Table 6.1 (4~ntiaued). External Storaqe Devices (VZO) 1. Magaetic Disk and lYlae~aetic ~Drtmm VZO (Direct A~ccess VZU) [4] VZU Capacity, Data Transmission 1'YPe of VZU C_e es Speed. iml?tes/s Interchangeable maqaetic disk YeS-5050 7.25 156 store ~ Interchanqeable ma~qnetic disk YaS-5061 29:17 312 store Interchanqeab~e magnetic disk YeS-5066 100 806 store Permanent maqaetic disk store Ye5-5060 0.8 150 Permanent maqnetic disk store YeS-5051 125 83.25 ' Maqnetic dr~aa store YeS-5033 6 1,250 Magnetic drum store YsS-5035 2 100 2. Magnetic Tape VZU Inforn?ation Recordinq Density, Maximwa Data Transmission VZU Code bits/mm eed, IOn?tes/s ~ YeS-5010 S/32 64 YeS-5014 63 126 ~ . YeS-5015 63 252 ~ YeS-5016 8/32 48 YeS-5019 8/22/32 96 YeS-5022 8/32 128 ~ YeS-5003 32/63 315 Buffer storaqe devices (BZU) are used to incri:ase the efficieacy of infoxmation exchange between internal and external storac~e devices havinq different speed. Buffer storage devices occupy an i.ntermedi:''ce Position in capacity and speed be- tween OzU and VZU. Magnetic drums aad disks are used as the information carriers in BZU. Multifunctional storaqe devices (MP'ZLJ), that alonq with inforaiation storaqe, read and wxite functions, also realize logic, arithmetic and special operations an the information stored in them,~have found application in third-generation oo~puters. The following operations caa be performed in I~'ZO: qeneral and simple associative retrieval, oode retrieval and comparison accorclinq to different nwaber criteria. minfa~um~ and mnximwa wo~d retrieval, closest mininum and m~xin?wa word retrievsl, ' retrieval of all n~mibers within qiven limits, any set of digit and complex lcgic - operations, arit2~metic operations and eo on. . The m~aory of the models of the tfiified Computer System is orgarlized on the hier- archical principle aad has the forna shown in Figure 6.4. 35 FOR .OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500060015-4 ruK ur'r~~~wL u~r. unLY , . ~l ~ Petuc~pa ~ HM,Q. ~5~ - v ~~NOR NaAp yHn~ (6) C~ . . /JQNA7b HM6 ~7 ~ 4 oNnre ~9~ ~ZO~ (~.l NMK t8~ = f-u y` pn6 e?-u" span~eae ~u~ He ~f-u yp~ Figure 6.4. Hierarchical Structure of Storaqe Devices of Onified Danputer system Key: 1. Registers 7. Maqnetic drum store 2. Processor 8. Maqnetic card store ' 3. Main storage 9. First level 4. Buffer storaqe 10. Second level 5. Maqnetic disk store 11. Third level 6. Magnetic tape store 12. Faurth level _ The local memory is at the first level. The local level of the YeS-1050 is desiqned on semiconductor flip-flops. The flip-flop registers have direct addressing and caa be used to store operaads, ad- dresses an8 indexes. The total capacity of the local memory of the YeS-1050 is - 864 bits and the cycle time is 160 nanoseoo~ds. T'he local memory of the YeS-1030 is an internal storage device with random access of addresses on thin cylindrical magsetic films. The capacity of the lxal memory is equal to 64 36-digit words and cycle time is d.6 aulcrosecands. ~ The local memory of the YeS-1020 is structurally contained in the main i.nternal storage (OOP)� Regardless of the capacity of the internal storage, that of the local menbry comprises 256 bytes. The cycle time is the same as that in the OOP. i.e., 2 microseoonds. The main internal starage (OOP) is located on the second level. It is desiqned on the basis of ferrite cores with rectangular hysteresis loop by the 2.5D system and has capacity of 64-1,024 Imytes with cycle of 1.25-2.U microseconds. Simultaaeous access to two internal storaqe units is possible in older mcdels, i.e., so-called double stratification of the menary, which ensures hiqher speed. The read-only memory used in the Unified Computer Systein to store control informa- tion and microprograms is also located at the second level. The third-level storage device includes a multiplex memory in which information that controls the operation of the multiplex channel is stored. It is an indepen- dent storage davice in older models of the Unified Computar System. The m~ry of 36 FOR OF'FICIAL USE 01~1LY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2447/02/09: CIA-RDP82-44850R444544464415-4 FOR OF~ICIAL USE ONLY the multiplex channel is desiqned in the YeS-1.030 on the basis of thin magnetic films, as is the locai mem~ry, and has a capacity of 1,024 32-bifi information words with minimum cycle time of 1 microsecond. The multiplex msur~ry in the YeS-1050 is based on ferrite cores with wirinq orqar.- ized by the 2D system. The menqry capacity is equal to 8 Kbytes. The multiplex mem~ry in the YeS-1020 is structurally co~tained fn the main internal storage. The capacity of the multiplex memory depends on the capacity of the OOP. - The fourth-level storage device consists of magnetic drum, magnetic disk, magnetic tape and maqnetic card stAres. The main complex in the YeS-1020 includes only mag- netic disk and magneti tape external storage devices. All types of external stor- age are provided in the basic complex in older models. Composition and designation of internal storage of YeS-1020 model. The internal " storage of the YeS-1020 model consists of three independent logic types of inemory: internal storaqe (OOP) , local memory (LP) and multiplex aienwry (MP) . A11 types of memory are structurally arranqed in a common memory unit. Access is possible only to one of the indicated types of inemary at each moment of time. Register (RNZ) is used as the information reqfster of all types of raemoiry and register (RMN) is used - as the address register. ~ao bytes of informatior~ are written or read in parallel code simultaneously upon access to the memory. The main internal storage with capacity of 32 K two-byte words is located in the same magnetic unit. However, the capacity can be increased by including additional magnetic units in the computer. Depending on the modification of the internal storaqe, the capacity of different types of the metaory corresponds to the data presented in Table 6.4. Table 6.4. Modification Capacity of Mem~ory, bytes � of OOP OOP LP MP YeS-3200-1 64R 256 76g YeS-3220-2 128K 256 1,792 YeS-3220-3 256K 256 1,792 The total access cycle consists of tw~o independent cycles: read and write cycle. The minimum cycle time of access to the internal storaqe, determined by the time between two sequential read instructions, comprises 2 micraseconds. Infonnation is read in the read cycle by the address stored in the RMN register, the read in- formation is recorded in the RNZ registsr and is regenerated in the memory unit. Information is erased in the write cycle by the address indicated in #he RMN and information is then recorded �rom the RNZ register. The total capacity of the internal storaqe is divided by the number of digits into 18 ferrite fields (16 infozmation digits and 2 verification digits). The ferrite field of each digit is represented in the form of a rectangular matrix consisting 37 FOR OF'FICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPR~VED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 r~sn vrr~..~h,. ~.aa, va.~.a , of 32 K cores (32 X 1,024). This matrix is divided into tw~o parts (sesaimatrix) measuring 512 X 32 each. T'he semimatrices are structurally arranqed on two sides of the plane of the magnetic unit ~MB) . A dia9ramof the maqnetic unit is pre- sented in Fiqure 6.9. ~ rTA ~ ~l~ ~((L/q-B(?36+4) BMXOQOB , B (2) ITR ~Qll/N-A(256+4)Bb/XOROB A (3) aW -re(az) ~ ~ - - ~ . 5 ~ ~ _ r~a - - - 18 - - 1` ~312+8)+32 ~ . ~ _o w _ ~ - N' ~ ~ ' ,QWP-f7~3?) . ~ . ~6~ ~ - C b iiP - - - _ . (s~z�a)�az ~ ~ ~ ~J ~ ' i b 0 ~ ~ i ~ ~ ~ . i i i i 4 . ,QU/P-1(32) � ~ ( ~ 1 � ~ . !TP _ _ _ = ; . 1 , (512+8)~3? ~ - - ~ - q ~ ~8~ !ldpecNae ~uuNe~ Ber3opxu[756�%) Figure 6.9. Diaqram of Magrietic Unit of YeS-1020 Caaputer Key: l. Address current geaerator 2. Address decoder-B (256 + 4) outputs 3. , Address decoder-A (256 + 4) outputs 4. Digit decoder 5. Diqit current generator 6. Code write bus 7. Diqit selection buses 8. A~ddress selection buses 38 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FOR OFFICIAL USE ONLY - A total of 512 coordinate buses on each side of the magnetic unit link the ferrite fields of all diqits in series and, beinq connected in pairs, form 256 address selection buses. Each address selection bus penetrates each digit of 64 cores, in each semimatrix of 32 cores with direct branch and 32 cores with reverse branch. The address selection buses of. each of sides A and B of the magnetic unit axe wn- nected to the address decoders DShA-A and DShA-B having 256 outputs each. The digit selection buses of each diq~.t are connected to their own digit decoders (on 32 outputs each). The shapinq of write currents in the decoders is controlled by the information being fed through the correspondinq code write buses (KShZ). The address (GTA) and digit (GTR) current qenerators excite the address and digit decoders, respectively. The selection semicurrent in any address bus acts in each digit on two cores con- nected by the same digit bus (cores F1 and F2 of the first digit). Dependinq on the direction of the semicurrents, the semicurrents in the diqit selection bus co- incide either in F1 in the direction of the first digit current or in F2 in the direction of the second digit current. Thus, the nvaaber of outputs of the address decoder is reduced by half on each side (256 instead of 512 are used) by chanqing the direction of the diqit selection semicurrents. Besides the 256 address selection buses of the internal storaqe, there are four additional address buses on each side of the magnetic unit that permit the re- quired capacity of the local (LP) and multiplex (I~) memory to be achieved. Each address bus permits a capacity of 64 two-byte w~rds, i.e., 128 bytes. Thus, the tw~o additional buses (one each on each side) provide a capacity of 256 bytes of local memory while the remaining six additional buses (three each on each side) provide a capacity of 768 bytes of the multiplex a~emory. ~ As noted above, the RNID1 address register contains 18 inforn?Ation diqits. The local memory address is lxated in digits 0-7 and the multiplex memory address is located in digits 0-10. The internal storage is constructed by the 2.SD type in older models of the Unified Computer System, the same as in the YeS-1020 madel, but differs in structural parts and main engineering and operatinq characteristics (see Table 6.1). The capacity of a single internal storaqe unit is equal to 32K 36-digit words in the Yes-1030 model, i.e., it comprises 128 IO~ytes. The cycle time is equal to 1.25 microseconds. The main complex of the internal storage contains two units with total capacity of 256 I~ytes. The internal storaqe in the YeS-1050 model has a capacity from 256 to 1,024 Imytes arranqed in units of 256 IO~ytes each. One unit comprises an individual device havinq total cycle time of 1.25 microaeconds. A word 8 bytes long can be selected simultaneously from a single memary unit. 6.4. High-Speed Memory (SOZU) As is known, modern computer systems have a hierarchical memary structure. The high-speed memory (SOZU) is at one of the first levels. A high-speed memory is . 39 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 h~R Uh'Hl(:IAL U~r. UNLY used to increase the productivity of nadern data processing systems during ex- ~ changea between the internal storaqe and the arithmetic-loqic unit or processor since the basic factor that li~aits the infoz~mation processinq speed is the speed of data exchanqe in the iaternal sttiraqe. Hawever, the frequency of access to the cells of the internal storaqe is usually verY is-re9ular. An ~itrarY access, oennected directly to a processar, permits the access time to those cells of the interna~ storsge in which the most frequently used data are stored (for ex- ~ ample, interca~sdiate results of calculations~ instructions that form the cyclic sec- tions of a proqram, soa~e service wards and ~ao on) to be reduced. A hiqh-speed mem4ry usually contains infoxmation used at the present moment or data which may be required in the veY'Y o~e~~eiredf data unitf fs+om the internalnstcr- SOZU upon access to it, rewrit ng r~ age is orqanized. The use of a high-speed n~enwrY is feasible according to the followinq concepts [16]: 1. High-speed mead~ries are used with specific ratio of access time to the main internal storaqe and speed of the arithmetic-logic device. 2. The hiqh-speed meinory in multiprocessor data processinq systems reduces the time losses related to switching the processors to the main internal storage. 3. The hiqh-speed mesnory in ~nioroPro9raun information processing systems can be used to store different user microprograms� 4. The use of a high-speed meaioi'Y in SYstems with multi,level microproqramminq leads to a significant increase of productivity. As already mentioned, the basic desiqnation of the high-speed mesnory is to increase the speed of the centra~ processor. It is natural tha~ the speed of fulfillinq programs will be higher, the fewer accesses are made to the maia internal storaqe. Therefore, information exchanqe between the hiqh-speed memory and the main internal storage may be reqarded as optimwa if it minimizes the total number of accesses to the internal storaqe durinq fulfilLaent of programs� Data is exchanqed between the main internal storage and the high-speed menpry usu- ally by the hardware method� T1ie charr.cteristic feature of the exchange is that both types of inea~ry are connected to each other and moreover each of these devices is connected directly to the central processor. If the infonnation required by the proqram to be fulfilled by the central processor is absent in the high-speed memory, the Prccessor either gains access to it to the main internal storaqe or it is immediately transferred from the iaain internal stor- aqe to the hiqh-sp~ed memory. The exchange is accomplished without interruption of the current proqram and without interference in the operating system since otherwise the operation of the central processor v~rould be slowed down. If the infora~ation to which the proqram qains access is not only the high-speed memory but is absent=in the main internal storaqe as well, fulfilL~ent of the current program is interrupted. 40 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 The structure of the high-speed memory depends to a oonsiderable deqree on the ad~opted method of i_.formation retrieval. The hiqh-speed mem~ory is divided by this feature into one with direct addressinq, stack addressinq aad with associative addressinq. The addresses of the hiqh-speed menary are indicated in explicit form in one with direct addressing in pmqrma instructions. Art example of this structure of a high-speed meupry is that of the coauron reqisters in mddels of the Onified Compu- ter System. Special four-digit addresses related to the high-speed mem~~ry are pro- vided in the instructions of the Unified Computer S~istem. The instructi~ format. detenaines beforehand which addresses are related to the main internal storaqe and which to the hiqh-speed memory and the proqram contains ccmplete instructfons on which data and for which addresses they are o~ntained in the high-speed memory. Direct addressing of the high-speed memory wilt be considered in m~re detail below on examples of specific models of the Onified Cbmputer Syste~n. The idea of stack structure of the hiqh-speed meapry is used in zero-address conr - puters in which there are no instructions in the qenerally accepted meaning. Each operand occupies a quite specific position in the mennrys therefore, there is no need to indicate its address in the instruction. Information exchnnqe between the internal storage and hiqh-speed meaary is controlled by special instructions that ~contain the address of the internal storaqe cell and the feature which indicates the direction of data transmission: internal storaqe-high-speEd memory, internal storaqe-processor, processo~internal storaqe and so on. The programs of problems solved in zero-address computers consist of an ordered sequence of operational and address syllables of identical digit capacity. The type of syllable is detez~mined by the feature diqit. If an address syllable is found in a proqram, the number is read from the main internal storaqe by a qiven address and it is referred to the stack. If an operation code syllable is found in the program, one or two nwabers (dependinq on the type of operation) fs selected from the stack, the operation is fulfilled and the result is recorded in the stack. The symbols representing an address or operatfon code in a proqram can follow each other or can be arranqed in arbitrary order. The working principle of a high-speed stack memory is an ordinary storaqe device ~ connected to a reverse counter, from which the access addresses are fed to the high-speed memory. The address of the first operand red from the hiqh-speed a~eabxy is determined by the initial s~catus of the counter, usually taken as equal to zero. Each time the next number is read from the stack, a one is subtracted from the con- tents of the counter, after which the requfred number is read from the high-speed memory by the address indicated by the cau~ter. This means that a sequential word file can be raad from a stack type high-speed memory only from a group of cells with sequentially decreasing nwnbers. The next new word in the high-speed memory is written by the address indicated by the counter and in this case the contents of the counter are increased by one. !tl FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407102/09: CIA-RDP82-00850R000500460015-4 ~ avx vrri~.~wa. ~JJl. Vl\l.1 The use of a hiqh-speed stack memory permits Practically complete euclusi,on of transmission of intermediate results betNeen the main internal stom9e ~d the ce~- tral processor. HoMever, this system is i~oaveaient Mhen orqaaizinq opezation in real-time aad time-shari.nq modes related to the need for interrup~i~ou, i.e., dis- ivption of the sequential fulfillment of a sinqle psnqram. lrbreover� stack orqaa- izatioA of a hiqh-speed mem~ry makes it difficult to use the ce11s of the stack memory as index reqisters, cycie oounters and reqisters for storaqe of retv~? ad- dresses after fulfillment of subroutines. The associative structure of the hiqh-speed memosY is used when or9aaizinq �~4e on demand betveen the high-speed meansy aad the maia interaal storaqe. i~ea de- siqninq the hiqh-speed meawry. the information of each cell of the me~wrY can re' place the infonoatioa of aay cell of th~ main ii?ters?al storaqe durinq fulfilLaent of the proqram. Each cell of the hiqh-speed mem~rY ccn~ains the inf�~ation field and address field. The address Which this cell w~oul.d have had if it belonqed to the internal storaqe is writtea in the adSress field. When an address is formed in the processor for the aext access to the memory. this address is ooanpared pri- marily to the oontents of the address fields of all the cells of the hiqh-speecl me~oory, if conforniity is found in aay cell, access fs qained to this cell and in the opposite case, i.e., if there is no oonformity in all the cells of the hiqh- speed mem4ry, access to the main internal storaqe is qained. The information field in each cell of the hiqh-speed memory is used to ~+rite and store the code oontained in the oell of the main internal storage beinq replaced� The strur.ture of the hiqh-spesd memory is called asso~fative because information is retrieved by the conforaity of the features, namely of the required addrass of the cell of the intes-aal storaqe and the oonteats of the address ffelds of the hiqh-speed memory cells, rather thaa directly bY the adaress of the high-speed memory cell. The advautaqe of a hiqh-speed associative aaemory is the proqram~fnq simPlicity, but its use is now restricted by the ccmplexity of enqineerinq realization. The hi,qh-speed meaiory in models of the tfiified Computer System fs frequently called local. Direct-addressing reqisters are used as the lxal memory in the YeS-1020. The lo- cal memory is co~ntained in the internal storaqe of the processor (see 6.2). The followinq are located in the local memory: 16 qeneral-purpose registers, 4 reqisters for floating-point operations, a current proqram status w~rd� range of channels, workinq meuqry of a processor and so on. The local menr~ry of the YeS-1022, like that in the YeS-1020, is used as cammon registers and operand reqisters with floating point to store the information re- quired durinq operation of selector and multiplex channels. The local memory of the YeS-1022 is an independent hiqh-speed memory oonstructed on sesaiconductor flip-flop storaqe elements in the inteqrated circuit version with arbitrary selec- tion. The capacity of the local memory is equal to 512 bYte$� The a�~ess time t� the memozy is 0.275-0.3 microseconds. ~l~ FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500060015-4 FOR OFFICIAL USE ONLY ' The local memory of tha YeS-1030 Mith raudan ad~ds'ess access has a~apacity of 64 r~rords of 36 bit~ sach in ~ach address and is ooastiructed aaa thin cY11n~lrical aag- netic films. Tbe storaqe elea~ent is the surfa~ce of a~+ir~ ooated ~+itb a thin filn ot a ferramaquet. The ~+~ozd Mindinq (read Mi:?ding) enoompasses a Mire, Mhi.le tbe wire itself (of beryllium bromae) is a Mrite bus and an outpnt ~?indin9 eimultan- . eously. The minim~aa access t3aN ot tbe local aemry of t2ie YeS-1030 is 0.6 ~aicrosacond. ~ The loval memory of the YeS-1050 is tande ia the fo~ ot reqist~rs based on 3i?te' qrated circuits r?ith random access. The capacity of the local a~m~ry is 96 bytas and access time 3s 0.24 micro~eoond. Chapter 9. Multiprocessor and Multimachina Co~puter b~st~s 9.1. General Data on Multiprocessor and llultimachiae Oo~puter 9yste~ns The tnain trend in developsen~ of coaputer equipment is further axpaas3roc? of the spheres of application of computers and as a o~ns~qt~eAC~ aaaversioa from sinqle machines of traditianal stzucture (processor-msawry-o~atrol-per3pheral de~vices) to oomputer systems and coa~lexes of diverse canfigurations rrith Mide ranqe of c~pabilities. Solution of such problems as development of larqe infonnation co~plexes, operatisiq in the collective-use ubae, desiqn of lsrqe oontrol systems caa~sistinq of a number of facilities aad operatinq in real time and processinq large iafoanation files places practicaily unlimited requirements on ooanputer equiPaaent and mnis~ly on speed and storaqe capacity. Ho~rever, the spesd of a sinqle-processor ccmputer reaches an order of 60 million operations per secoad [6], i.e., closs to the physical limit, at the modern level of development of se~ioonductor integrated circuit technology. Therefore, a further increase of the speed of caoputer equiPment is possible oaly by new structural ~solutions qualitatiwly different from traditioaal solutioas. One of the methods of providinq more effective joint operstion af devices cantained ia coa~puters is multiproqraa~ainq. Th~ n?ultiproqram~ainq principle of operation re- quired a significant chanqe ia the structure of sevo~d-qenerstioa oomputers. De- ve lopment of multiproqrammi.aiq not orily permitted an iacrease of the real productiv- ity of compu~ers, but also made it possible to lay the qronndaork for the appear- ance of machines oriented totirard operatinq ia tha time-sharinq mode. Florar sheet-proqram devices that ensure reliability of multiproqram opsration, for example, proqram interrupt and priority floa sheet, memarl? protection f1oM shset, ~ relative addressinq flow sheet, CONTROLLBR pro9ram Permsnently in the computer memory arid so on, were in:.~luded in tha computer structure. l~ further desirs to increase the productivity of camputers led to the development of machines that included sev~eral devic~s that perforn~ tha same function (for ex- ample, data storaqe, fulfillment of arithmetic operations and so on). In other . words, duplfcation of different devices (caatral processors, storaqe devices, in- put-output processors and so on) appeared in the computer structure. 43 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500064415-4 ruK urr~~iw~ ua~ vr~Y Duplication of irxlividual vaits of the computer aad also the presence of special control proqra~ns that Permit pnrallel fulfillment of sepnrate operati~ons of the computer process ia differeat devices aigaificantly increase the productivity of the oomputer due tt~ msxinnm? lcadin4 of devices and the reliability of the co~uter aystem is enhanced in this case. A computer system may include different devices with ideatical functions (for ex- a~ple, those that perform arithmetic operations), autonomous devices of the same type (several parallel models of internal storaqe) or a liierarchical complex of devices (one caatral and sev~eral auxiliary process~ars). A system which oontains tw~ or several inforn~tion processinq devices. i.e.. Pro- ~essors fuactioniaq imder unified oontrol, is usually called a multiprocessor system. parallel operatfon of several processors that simultaneously fulfill sev- erai pmqrams or different pnrts of a single larqe proqram are realized in multi- processor systems. Alonq with multiprocessor coa~puter syste~as, multimachine computer systems (com- plexes) are used. Conguter systems containinq two or several identical or different independent ma- ch ines connected to each ~ther ttirou9h an excl~anqe device are called multimachine computer systems. Hoth m~ultiprocessor and multimachine computer com~ple~oes are oanbii?ed into a broader class--computer systems havin4 9eneral features. 2~ultiprocessor and�multi- machine conQuter syste~ns are now beinq developed on the basis of highly productive computers and are the basis for a network of collective-~use camputer centers. Computer systems permit on the one hand a considerable increase of caanputer equiP- ment productivity by parallel operation of devices ~~or example, of a proc.esgor in a multiprocessor syst~a) or of sev~eral computgrs contained in a multimachine coan- plex, and on the other hand they considerably facilitate man machine interaction, providinq a broad capability for direct himman interference in the proqram fulfill- ment process. The main requirement on computers used in con~puter centers is to pirovide the capa- bility of their operatinq in the collective-use mode, hiqh reliability, the capa- bility of i.ncreasinq computinq capacity, adpatability to chanqes in facility con- tYVl functions and so on. Camputer systems, to meet these requirea~nts, should: have s developed operating system that ensures simultaneous fulfillment of different programs and that provides access to standard subroutines by users; have translators from algorithmic languages to facilitate the proqraanner's w~ork in prcgram preparationj contain devices that ensure dynamic distribution of inemory a~aon4 Pro9r~ ~d also free movement of proqrams during calculationst ~1 F'OR OFFICUL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500460015-4 FOR OFF[CUL USE ONLY have memcry and proqram protection devices fs~om interference of ather ps~o9rams t have a time sensor (timer) that permits allacation of tbe required time to them acoordinq to the users' inquiries for ~ork, durinq Nhich the systeu is switchsd autamatically to fulfill other proqramst have both hardware and softWare to orqanize priorities for s3multaneously waitiu9 Pm9rams. One of the basic featslxes of oo~oputer systens is the extre~ely large diversity of their structures. Their classification baco~mes neceasary in this regard. Let us oonsider the different approaches to classification of ooaiputer syaten?s. Accordinq to the principles established durinq desiqn of casnputer systems and acc~rdinq to the main fields of application, vo~puter systems are divided into three qroups: . 1. Computer netw~rks r?hich are developsd vn the basis of existinq oomputer models by combininq tsrritorially dispersed oo~oputer centars by means of data transmission apparatus and communicatior?s chsanels. An example of the qiven . qroup msy be the State oomputer Ceuter Netw~ork (GSVTs). Computer systems used in computer networks are oousidered in detail in the third sectia~ of this textbook. 2. Computinq media ahich are coanputer systems oonsistinq of a larqe nuaober of intercosnected componeats, each of which can be proqram adjusted for informa- tion processinq, storaqe and transmission. The use of microprocessors as one of the desiqn elements of computer systea?s and also the dsvelopmeat of miniature hiqh-speed and hiqh-capacity mem~ries are of qreat interest. All the co~ponents in con~uter media operate in parallel and can simultaueously realize a n~nber af calculatinq processes. The productivity of computinq media may increase essen- tially without limit with a sufficieat nwnber of compoaeats with~out a significant - increase of the oost of calculations, i.e., the use of computinq msdia may result in a larqe saviaq. However, computinq media have not yet fouad sufficiently broad application. ~ 3. Multiprocessor aad multimachine computer systems (comple~oes) o~mbine'the two basic trends in development of canputer techaology--an incrsase of reliabil- ity and capability of recoafiquration (flexibility). This is achieved due to the mcdular nature of design. Let us consider the characteristic features of designinq this qroup of Multipmcessor and multimachine computer systems (or complexes) can be classified by different features: desiqnation, type of equipmeat, method of controlling individual vomponents of the ~ystem, permar?ence of structure, degree of terri- torial dispersion and so on. . The suqqested classification of multiprocessor and multimachine computer syste~as is presented in Fiqure 9.1. 45 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED F~R RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 r~w ve~n.~~+a. .,oc v~~a.� � ~ . ; w~rucn~u~tnw~we ~ dJCTCMbI ' IZ) ~ yBC ~ . : Maa~Y+~oBHMe ~ A~obw~NNoro Co6NeweM~rczv runo r~no . E4~ ~10)~ . C nrpeHeNMOU C noc or~ crpyKrypou 5 c~pyK''IP�~ L~CNT/JO/1UJO~ON- CL(eNTpGAUJO� 0 CMtI(/GNNA/M N!/C ~ ~QNMA~C yn/lQ~IItNI/CN C ,,.'~CCCTKlIN ~ . ,IU?Q~QIOl1(UM " loxpenneNUUV aaxpenneNu ~,yN~ctuu ~ q~yn~tuu Figure 9.1. Classification of cbmputer Systems Key: 1. Ccmputer systems 8. Specialized computer sYstem 2. iJniversal conputer system 9� I~OmOg~~s 3. ~omoqeneotis 10. Combined type 4. Dispersed type 11. With permanent stru~turg 5. With variable struc~ure 12. Decsntralized 6. Centralized 13. with mixed cantml 7. With "riqid" securinq of 14. With "floatinq" sacurinq of functions function Caaputer systems can be universal (WS) and specialized (SVS) by desiqnation. Universal com~uter systeoas are desiqned to s~lve a wide ranqe of problen~s. Spe- cialized con~uter systems are orianted toward ~olutioa of one or aeveral problems of the sam~e type� The rariqe of problems designe8 for solution by specfalized computer systems usually envisions hardware and sof~s~are desigsed specially for this system. Thus~ for example, specialized computer systems that cantrol produc- tion processes should contai.a aaalc9'diqital ooaiver~srs a~d vice versa. Devices for displayinq the results of information procesainq special~y developed for them� 1~6 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 F~OR OFt~7CiAL 1 for example, illuminated display boards in operatio~al control of production, qraph plotters aad so cs?, are used extsasively in specialized co~puter syste~s. Computer systems are divided into homoqeneous aad inhom~geaeous bY tYPe of equiP- ment. Homoqeneous systems ooatain sev~ral processors of the same type ~i.ncternsl storaqe nadules aad so oa) or caasist of sev~eral proqraarcampatible aschines. In homo qeneous con~puter systems usually iaclude devices of different typa or dif- ferent ooaQuters. T'he use of homoqeneous computer systems is especially effectivs in larqe hierarch- ical complexes. The presence of equipment of the same type at all oontrol levels pernits the use of a uaified lanquaqe and ~mffied software, whf le the modular principle of eiesign considerably simplifies maintenance aad reaesiqn of the syste~ to increase its capacity. It is m~re conveaiernt to distribute the Mork bet~een users in homoqeneous systems. The disadvantaqes of homoqeneous systems include incomplete utilizatica of the productivity of all computers (processars) coutained ia the system. The effi- ciency of wsinq them depends on the deqree of loadiaq of individual vomputers (processors). Inhomogeneous cvmputer systems caa be used to increase the utilization efficieacy of oomputers with differeat productivity. For example, a more product~v~e coaputer processes inforaation vrhile a less productive computer is used to eater aad re- _ trieve information (based ~ the BESM-6 and MIR). Special translators and equipment that ensure integratioa of the devices or indi- vidual computers contained in the system are required for normal operation of in- - homoqeneous systems. Computer networks can be divided iato tMO types--cvml~ined and disPersed--acoordinq to the deqree of territorial dispersion. Combined systems include those in which the information traasmissio~ time from one computer (pmcessor) to another (another processor) is neqliqible coa~pared to the time required to prxess it on aeie of the machines. InfoYmatio~ is trans- mitted over communications lines from one device (coaaputer) to aaother usually in parallel code. Multip~ocessor systems are usvally of the caabined type. The information transmission time over com~nunications chaiusels betWeen oomputers in dispersed systems is comparable to the problem-solvinq time on one of the ma- chines and therefore it should be taker? into account when investiqatinq the fuac- tioning of the system. Information is usually tranemitted in seqneatial.code in this case. Computer systems are divfded by permanesice of structure into systems with constant and variable structure. The structure of a coaaputer system is understood as its composition and connection between its components. Tw~o oomputers contained in a system are reqarded as functionally oonnected if transmission of functional inforaation (proqrams, input and intermediate data of 47 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 rv~c v~ri~~w~ uaa: vi.~i problems being solved) fro~m ane mschine to another is possible durinq their operatioa. ~ 1~ computer systam is called oriented accordinq to functional c~a~aections if the ~unctional infoYmation can be transmitted in only one direction. A computer sys- tem is reqarded as nonroriented when information is transmittad in both directi�A~. A aomputer syste~n can also be partially oriented, i.e., it can include both ori- ented and nonrorisnted subsystesos si,multaneously. Oomputers cacitained in a system can be oonaected by the contml principle: can- trol infoYmation is fed throuqh one oomputer (co~trol oomPuter) to another (con- trolled). The system is called directional if the co~trol c~naunications are oonstantly directed in one directioa. The c~ontrol communications can be directed in both directions in a nondirectional system. Both combined and separate coao~unications channels can be used to transmit fvac- tional and oontrol infoxmation. A system iu which the composition of functional and coatrol coaimuaicaticns does not chaaqe durinq functioninq is called a ca~aputer system with peYmanetit structure� Adaptive systmns, i.e., systems in which the structure changes cn the basis of aaalyzinq curreat int~rnaatioa, have variable structure. These systems permit one to achieve an optimum state uader aay variable functioning aouditions. A computer net~rork can be divided into ceatralized, decentralized and mi~oed von- trol accordinq to the deqree of oontrol centralization. A typical fe~ature of a centralized system is the presence of a central control coa~uter (processor), called a director-computer. A director-camputer controls the functions and exchanqes information in a system and coordinates the operation of all machines of the system. Moreover. it can be used for calculations. The advantaqe of a centralized system is that optimian methods of problem-solving can be found during its fuactioninq. When solvinq complex problems, the director- machine distributes the work amonq individual computers (Processors), thus Provid- inq a reduction of the total time required for calculations. Each computer (processor) in a decentralized system operates autonomously and all machines solve their oam parts of pmblems. A decentralized system can have a coaanutatinq device throuqh which infoYmatia~? is exchar~qed in the system. Both centralized and decentralized contr~ol principles are realized in systems with mixed ooatrol. The entire computer complex in thes~ systems is divided into groups of intexactinq computer modules with centralized oontrol within each qroup. The considered classification does not claim to be complete and can be supplemented with a number of other classification features (for examPle, in time mode, operat- inq priaciple of the computer and so on). The followinq main indicators--productivity, econon~y and reliability--can be used to characterize and analyze coaiputer systems. 48 F'OR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500460015-4 9.2. Desiqn Principles of Multimachine Computer Systems (Complexes) Durinq the initial phase of their developme~� multimachine oomPuter complexea (1~lVK) were used to v~erify the correctness of equfp~aent operation and also to in- crsase the reliability of the results obtained. Two identical machines simultaneouslY Performed the same task with the same data. Intera~ediate and final results were compared to each other. l~breoner, so-called duglex coaputer systems arere used which consisted of t~ro individual sets of can- puter equipment, each of which could take csi itself the functions of the system and in this case the sacond set could act as a etandby. Thus, one of the computers operated as an active unit and the other was a standby unit. The machines could interchanqe roles automatically or semi-sutomatically during operatfon. Processinq was not duplicated in these syatems. Part of the data required to ccntinwe calculations could be transmitted psriodic- ally from an active ca~aputer to the standby computer through a buffer storaqe (for example, maqnetic drum storaqe connected to both machines). Thus, the main data were store8 in duplicate tables of the durinq problem-solv- inq. This provided the capability at any moment of time of transferring the active mle to the standby computer, which could continue the calculating process of this san~e task. Multimachine oomputer o~mple~aes can be dividefl by this type of orqanization into two groupss unconnected and connected canputer systems. Unconnected multimachine oomputer complexes were developed to relieve the central processor of perforn~inq data fnput-output operations externally. ~ They consist of central and peripheral computers between which there is no direct physical vonnection (Figure 9.2). There is no jointly used hardware in unconnec- ted multimachine canputer complexes. The feas{bility of usinq them is determined by the fact that input-output and calculation operations are combined in time. A small and inexpensive machine performa slow information input-output operations (readinq from punch cards, printout and so on), while a central computer performs hiqh-speed operations, exchanqing data with external storaqe devices during calculations. Practical ~lementation of these~ operations can be shown isi the followinq exam- ple. Information is transferred from puncl'i cards to maqnetic tape, where the as- signment pack is formed. A peripheral machine controls these operations. The maqnetic tape store is conneated to a central computer in which calculating oper- ations are performed. The magnetic tape store servicing a single machine is switched to service another machine by a special auton?atic switch. Data output is controlled by a peripheral con~uter to which the magnetic tape store is con- nected durinq this time. The multimachine computer complex has hiqh productivity in this case due to the zrore efficient use of the central computer. But the total infornation processing time, including infoYmation printout, is increased for each user with this mode Lt9 . FOR OFF'ICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 rux urr~~iw~ wc ~rLY . ~ ~ ~ :w co oc o~ ~ C bq � p~m y"',~ �o dl 'q i ~ ~a ~ ~ ~~o ~x U C3 a v ~e~0 ~ ~ '~N ~~i~'? ~.o =,.no.~ ~ 4~ C�xa~�~ f1~ x . ~ m a a~ ~ a ~ i== a q ~ ~ ~?r . 4~~ V ~ ~ 3 ~ w 0 `o~�~ p~ i~ ~ uyo v C a: ~ ~ Z x ~p . ~oo o ~ ~ d a b ~ ~ � ~ ~ A1 m ~ ~ a1 ~C j ' ~ ~4 b ~ i~i ~ ii F i~ ~n a ~ ~ 0 o u � 8 a ~ ~v 3 a ~d O b ~ O O.~ I x V m ~ 1~ d 41 ~ ~i ~3 ~.j~3 O+a1 8 ~a8a �i~~ ~m ~ mw~b~~~~~'p'wa . ~m�w~' o~ �0 a.-~~'~ o+g~ ~C N dl ~'d ~ i~ 41 O ~ . +~~r}~ama~s~m~er~~ - ~ �~N ~~fa~ ~ ~ ~ a m u~+ ~ w v+~ :a�`~~~'~ wt~3~a~v~~~~~~i~a C~ ~V.. ~ ~ C ~ J~'V � � � � � � � � � � � � � W rl N en d~ v1 ~G r CD d? O.-1 cV rf rl .-1 rl .-1 ~ ~ 50 - FOR OFFICIAL USE ONLY / APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007142/09: CIA-RDP82-40854R040500060015-4 rvn va� v~.. Vl\Ll .i ~ ' ~ .y ~i � . f..: . . v v �2 M6 ~ {5~ t . , ' � . Bayucnu- ~ �Cucre~a ' � � . reneNaa : , . deo8a- sei- ~3~ueu � , ' � , CuCTQMQ ,t~yr~ . eoaa . . ~ :;}f . . ~ ~ . ~4} Mn � � ~ � ; , ~ ~ ::~,,~.;;.'1; :~'kF'.e: Ai F'igure 9.4. Indirectly or Weakly Connected System Key: 1. Input-output system 4. Magnetic tape 2. Magnetic disk or magnetic driua 5. Computer system 3. Or of infonaatian transmission over input-output channels, the'communication of one machine with another through a channel adapter requires matched oFerAtion and a specific sequence of actions of all computers. When one machine has access to another, a preliminary messaqe must be transmitted to it through an adapter. T'he adapter perfozms the role of control device. It receives the instructions from the oommunications channel and decodes them, providinq connection of the channels = and synchronization of operations performed by both channels accordinq to these instructions. Both machines sheuld set the corresponding information read and write programs to working status due to the action of iaquiry instructions and the adapter primarily authorizes data transmission. Multimachine con~uter complexes began to be developed both abroad and in the USSR on the basis of first-generation computers. The first multimachine co4nputer complex that became known (the Sage antiaireraft defense sytem, United States, 1955) was a system consistinq of 15 computer centers connected by coa~aunications lines, in each of which two computers of the same type w~ere used and one of them was a standby compt:cer. The main task of this canplex was to ensure hiqh reliability of the compe'cer system. Homogeneous type multimachine compu~er complexes have been developed in our ooun- try: the M-222 (based on the Minsk-22 coirmmuter), the Astra�(based on the Minsk- 32 com~uter), the t4i~imnx (based on the M-6000 computer) and the Sunana (based on - the Elektmnika-100 computer)= inhomogEneous multimaahine computer complexes have been developed on the basis of the BESM-6, Mir-2, Aist (based on the M-20 and Minsk-22 computers) and other co~~puters. The M-222 multimachinP computer complex is a universal homoqeneous computer system _ with variable structure. 52 . FOR OFF[CIAL USE ONLY ~ APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 F'OR OFF7CIAL USE ONLY since it must wait for the com~letion of data ccaversion p.rocedures ut the input - - and output of the system for the entire pack. Co~nected multimachine oomputer coa~lexes include ~ev~eral machines which jointly utilize coamon hardware, i.e., electric inteqrafiion between processors is possible in these systems. Both machines fn a connected multimachine camputer complex can fulfill t~ro differ- ent programs autononausly or in interaction with each other. Interaction betWeen ~ computers remains at the data exchange level since the main computer utilizes the other machine only as an input-output device. Indirect (weak) or direct contact is pcssible between co~nputers. Direct-connect multimachine computer comiplexes have rigid electrical cannection - between machines at the level of a conman hiqh-speed mem4ry or by direct.connec- tion of tw~o high-speed channels. These systems include tw~ or more computers, one of which is a highly productive machine and the remaininq ones are used for infonaation input-output and preliminary processing of it (Fiqure 9.3). Co6MecrNO ~ Cucrer?a ucnone~yeHaa Bayucnu=� GGodb- 6o~crpoBeucr- ~eauwn . � ~l~ Basoda ~~~yvA ~3~ cucreNa.'~~ ~ ~ . noHr+ra f 2~ ~ 1 . ~ 1 Ba~vac/w- . t'uc~eaa ,i.Q. t4~ qDanrep~K~Man ~ reneNOn ~ deodh- _ucnMp ~ QaQoda . ~ , , . , Figure 9.3. Direct-Connected System Key : l. Input-output system 4. Channel 2. Jointly used high-speed memory 5. Adapter channer 3. Computer system In weakly connected systems, the computers have coannon use of part of the input- output equipment (for exmnple, magnetic disk or maqnetic tape external storaqe). Interaction between the proqrams being fulfilled in each individual machine is in~ossible. Exchanqe is carried out only at the information level. T'he first oomputer enters data into a common storaqe, whfle the second machine can have ac- cess to this storaqe device at its own initiative or on a siqnal received fro~n the first computer (Figure 9.4). Thus, systems with weak connection incl+~de such multimachine comput~r c~mplexes in which the role of work distributor among all the remaininq computers engaqed in task processing is allocated to a single computer. Computers are usually connec- ted to each other by means of a channel adapter--the channel. Unlike other forms 51 .FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~ ' Productivity can be increased if necessary by simple c~nnection of auxilfary coa~puters. The M-222 multiuuschine oomputer caaplex is a one-dimensional system with tw~o-way coanaunication channels betwaen elementary vomputers. The nuznber of computers in tha syatem can bis varied in the range from 1 to 16. Each computer consist:s of the Minsk-22 oomputer and a systems device which includes the operat- ing syst~n unit a~d oondntm~cation channel ooamutator. Ths structure ~f a multimachine caaputer canplex can be chanqed to isolated sub- systems by means of the tuninq reqister avsilable in the system. The ornamunication channel cammutator can open or clos~ conmunications channels depen~ding on the status of the tuning reqister. The main features of the 1~222 multimachfns computer complex oonsist in the followinq: 1) hom~geneity. All machines are identica~.. ~'here is no direcbor-computer. Any computer can take a~ its fusctionst 2) an n-fold increase of productivity is achiev~ed by simultaneous perfos~naance of operations on all n taachiaes of the systeaa= 3) variable structure. T'he system can be divided by program into subsystems and machines can be noted in the subsystem that jaintly perform exchar~ge, condi- tional and unconditional trauisfer operations. This method permits one to alter the structure of the system as a function of the task beinq solved a~nd significant- - ly enhances the structural reliability of the complexj 4) an increase of the system by simple connection= 5) the complex is synchronized by the proqram method. T'he operation of all vomputers of the complex is synchronized by the qenerator of a single (any) ct~mpu- ter of the complex at moments of informatfon exchanqe. The Minimax multimachine computer complex is related to those with program-switched comaaunications between individual computers. The complex is constructed on the basis of ASVT (modular computer equipment) hardware and software. Every comPuter of the multimachine ca~puter complex can have its own individual external devices and moreover can have access to allocated external devices called system devices. This complex'can be used in ASU. Z'he SumQaa (minicomputer contro~ systean) multimachine computer complex is related to homoqeneous ccmplexes (it has a proqramaable structure), has the capability of increasinq its capacity over a broad ranqe, is characterized by high reliability a,nd is constructed on the basis of Soviet Elektronika-100 and Elektronika-100I tainicomputers. The S~unma multimachine ccmputer complex consists of computers of the san?e type. ~ao-directional proqram-switched communications channels are used to exchange control inforn?ation and data bet~a~een computers of the multimachine computer complex. The multimachine computer complex includes Elektronika-100 - computers and a systems device that pe maits the structure of the multia?achine computer com~lex to be programmed and that permits systems interactions of the tnachines. 53 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500460015-4 rvn vrr~a.~~?.. voa, v~.a.~ The systen~s device is co~ected to one computer throuqh atandard chanaels for ex- ternal devices and to the systems devices of four ather coni~uters throuqh inter- machine oomaaunications channels. ~fa~o functional modes are most frequently realized #.n the Swdma n~ultimachine compu- ter vomplex: one modes is used when solving scientific and techaical and fafoYma~ tion problems and is the collective-use mode aad the other. is oriented tvward real- time servicinq of the priority flow of declarations. , Development of multimachine coanplexes based on models of the Onified domputer Sys- tem is of special interest. The capability of combiniaq several aachines into a Unified eomputer System is embodied both in the structure aad software of the Unified Canputer system. Conmaunications between individual computers can be accomplished at the lev~el of any of independent, loqically independent devices (processors, internal storaqe. multiplex and selector channels and so on) by means of hardware and software. Dependinq on the type of hardware used, the tollowing levela oE commuaications between computers are distinquished: 1. At the cha~nnel level by means of the "channel-chsnnel" adapter. The "channel-channel" adapter (the YeS-4060 device) is used to exchanqe data between input-output channels of the Unified Qomputer System and transmits infonaation from the internal storaqe of one computer to another throuqh the channels of these machines. The adapter can join mul~iplex and selector channels and also t~vo se- lector channels. It operates only in the monopole m~de and transmits data at the ~peed of the slowly operatinq channel (of the two co~nected channels). It per- forms the functions of external device control, which is selected by the channel, fo r each of the channels to which the adaptex is ca~nected. The adapter answers inquiries of the channel and receives and decodes instructions of the channel. The adapter differs from any other external device control by the fact that it does not control inputroutput devices but only provides ocmmunication betweer? chan- nels and synchronizes their operation. T'he �channel-channel" adapter consists of two contml units, each of which is con- nected to an inputwutput interface, and services its own channel. The control units are oonnected directly to each other by several sicn?al lines and oomawn single-byte buffer reqister to which the information transmitted from one channel to another is fed. 2. At the level of external device controls by means of a tw~o-channel switch that periaits w~orking, for example, with two channels belonginq to different machines. 3. At the level of the external meiaory or input-output devices by means of a two-way switch. Connnunication at this level is feasible if hiqh speed is not re- - quired. This method provides a considerable increase of the v~ol~anes of program and nwnerical information simultaneously accessible to pmcessors in multiprogram operation. The external device controls comm~unicate with the two channels throuqh a standard input-output interface. The external device control operates with the channel which first sent the inquiry to work with an external storaqe device. 51~ FOR OF'F[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FOR OFFICIAL USE ONLY 4. At the level of data transmission equipnient by means of auxiliary remote communications control devices. 5. At the level of a con~non internal storage by means of the ccnsole that switches the meawry units. This oommnunicatfon permits one to orqanfze c~ontrol of the calculatinq process of a vaified operatinq system aad to achieve simultaneous solution of independeat parts of the same problem. Parallel access of the proces- sors to the memory uaits is suthorized due to the double ooamuaicatioas interface of the memory uaits with the processors and priority circuit. The reocnfiguratic~ - co~nsole oonnected to the system permits one or several n~enqrY units to be excluded and also permits distribution of the aiemory units amonq processors- 6. At the processor level by means of direct control devices. In this case siqnals of external interruptions and wntrol information is transmitted bet~reea processors to synchronize the unified calculatinq process. Multimachine computer systems based on older models of the Onified Oomputer System, for example, the YeS-1050 can be constructed by usinq any conmaunications level, but the functional efficiency of the oomputer complex will be different. Younqer models of the Unified computer complex can be coanected only at the channel or common external device level. - The main task of orqanizinq communications between machines at any level, as is known, includes the capability of usinq coarnon isformation files and proqrams by connected processors of the system. The mcst flexible and hiqhest speed is com- munication of processors through a coma~n internal storaqe field. In this case the calculatinq capability of the system and the utilization efficiency of the in- ternal storaqe are increased sharply. Mcreover, the calculatinq system can control a unified operating system. Based on models of the Unified Computer System, calculatinq complexes of two ma- chines are primarily developed. A diaqram of a two-prxessor computer system that includes different types of ooamunications is presented in Fiqure 9.5. Specific- ally, the VK-1010 multimachine computer complex has been orqanized on the bnsis of two Yes-1030 computers connected by direct control buses to the computer c.wmplex status unit (BSVK). Tiie VK-1010 computer complex can aperate in five different m~des: ' l. The computer complex is in working order and operates in the "hot" stana~by mode. In this case one of the machines is the main one and the other is a res~erve. Poth the main and reserve machines receive information coming from external devices and process it by the same algorithm, but the information is se~~t L-o external devices only by the main computer. If the main c~oaaputer is un- s},Ye to transmit the infonaation for any reason, the reserve computer takes on the functions of the main one, leavinq the main wmputer as the reserve. The YeS-1030 computers are switched to the MAIl4 and RESERVE modes either by proqram or by an operator from a disnatcher's control console. 2. One of the machines is in working order and solves problems without dupli- cation and the second computer is also in workinq order but is in the preventive 55 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 r~l~c tN~r~a,rwa, u~a vi~a.a ~ ~ 6/?OK ~ ~IAGMI � 6QbK ~ 6~4aK ~ L~~ J ~ n .,1 L~ ~ ~ n nAar~ecrno Nnre eaC oto y~pa neNU~ t-- r -r--t ~ ~ ~ Adbnrep r 1 I ( KaNan I llco~b~ Kcieon i koNan ' - L---~ L---J NNre c f--~--- --=--T---f r-1--~ t~~ r-l- ~ i !ISIBB ; ~IyBB SIlIBB i Sl6/BB ' ~----J L---~ f- -r - - - - - --t r_l_~ ~_-1_~ - i By ~ ~ B!I ~ By ~ L---~ L---~ gigure 9.5. Diaqram of ~o-Ps~cessor Camputer S~?stem ~Y = 1. Main internal storaqe unit 5. Chanael-chaanel adapter 2. Processor 6. Input-output interface ~ 3. Direct vontrol interface 7. ~put-output control device 4. Channel 8. External device maintenance mode durinq this time. If the main machine malfunctions, the preven- tive maintenance w~rk on the second machine cari be stopped and it caa be sv?itched to solve the basic proqram. The momeat that the main canputer is saiitched off can not be predicted beforehand since it is determined by the resultinq situation. 3. ~e of the machines is inoperable. The problem is solved on the fvaction- inq computer while the inoperable coa~puter is checked by mafatensnce tests. if a . malfunction is detectsd, the cmmputer is switched to the repair mode by the opera- tor from the camputer complex control ccnsole. If a malfut?ction is no~ detected, then a breakdown in the computer is recorded and the machiae continues in the pre- ventive mafntenance state. 4. One of the machir~e~ is in repair. After canpietion of repaire the computer is transferred to tha preventive mainte~ce mode to de~es~mine the quality of repair. 56 FOR O~FICUIL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 . 5. ~u auto~~us mde of cparation of mchines contaisicd in the oo~oplsx. Each oo~patar oparates nith its oMa sat o~ ~sternal devic~s in~apa~entlp of th~ other. T!w statea of oo~put~sa ooa~s3,d~red above are shoyn in Table 9.1. ~ Table 9.1. YeS-1030 (1) Y~8-1030 (2) Oti, Fnactioninq. Maia 0li, Functiauinq. R~ser�ro . Fuactioeiinq, I+lain Plmctia~?inq. Psyv~ent~ve Maintenance Fvu~ct3,oniag, !lain !laltnacti~oainq~ Preventive l~lainteaauce Fuuctioaiaq, IKain lyalfuactioainq� ~ ~P~r Functicainq, IndeptAdent Functicstinq. Ind~Peeideut - Tha operatinq mode of the c.carputsr caoplex can be assigned by p~qram or by dirsct oontrol instructions or (d~pendinq on the daveloped sitnation) fs+o~a the caoplex vontrol oouaole. The VR-1010 com~uter complex providea: orqanization of a ca~on internal storage lield, i.e., independeat access on the part of each of t~a processors throuqh t~a maialine syst~ms and the capability of simultaaeous access of each of the p~ocassors to differeat intexaal storaqe nnits thr~uqh a caamon iaternal strorage unit= exchanqe of coatrol siqnals of computer status for automatic svritchinq of the operatinq ta~des of the oo~puterst e~ochanqe of control informatioz~ batMSen tha processors of tMO YsS-1030 comp~ ters of the complex through direct caatrol channels= exchanqe of information between t~ro mnchines by means of the "chaauel-channel" adapter; ooz~nection of external storage dsvices to any of t~ machines. A complex of tvro YeS-1030 machines usir?q mit?icomputers as buffer processors has cansiderably qreater capnbilities. Minic~mputers take oa themselves the functioas of the lo++ distribution of users aad are communicat3,on processors. Creation cf a two-processor syste~n pith coanan iaternal storaqe field is possible on the basis of YeS-1030 and YeS-1050 mode.'_~ or of trro YsS-1050 mnchinss. The system includes a rec~fiquration oonsol~ to assiqa the operatinq models of the complex, physical distribution of system resources (internal atoraqe, external device contsnls and the exteraal devices themselves) betaeen processors and to exclude malfunctioninq devices from the system. A t~ro-procassor systam has direct control devices that include auxiliary lines ~fiich are used in specific operatiaq aades to synchx+o~ize tha ~ork of processors and to transmit control infortmtia~?. 57 - ~OR O~'F'[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 rvx ~rr~a,~w~ u~a ~n~Y The processors of a tw~o-processor system can operate in three nades: MOLTISYSTEM, MDDEL and SPLIT SYSTEM. In the I~IUI.TISYSTBM mode, two psocessors are combined into a wnified system with internal and external storaqe split betNeeu them and also aith input-output dv- vices to co~trol the coaroon multipmcessor control proqram. Independent operatian of the computer organized by a sinqle-processor coc?trol pro- qram is acoompliahed in the 1rlODEL mode. Multisystem devices are not used aad mul- tiaystesn siqnals are not emitted and are not received. In the SPLIT SYSTFi~I mode, the multiprocessor control psbqram is carried out by us- inq a sinqle processar, but it has access to any of tw~o psbcessors. This m~de can be used if one of the processors mslfunctions. Zhe action of direct control in- structions is blocked in this case with communication at the processor level. Multisystem instructions are not issued and are not received~since the processors operate indepesidently. ocmversion from the MODEL to the SPLIT SYSTEM mode is made from the system reconfiquration cansole. Work in the modes described above is organized by memns of a developed operating system that includes a contml proqs~m which ensures multiproqram~ainq with varia- ble nwnber of tasks. The characteristic feature of the operatinq system is the use of processors as a unified resource when processinq the assignment priority, perfoYmiaq input-output operations for the task solved by a single processor usitiq snother and Processing error siqnals in an apparatus by usinq tw~n processors. The required devices to orqanize a time-sharinq awde are provided in older models of YeS con~uters (for example, the YeS-1050) in which a service period by the processor is made available to each independent task sequentially (in the order of - priority aad sequesce) with svbsequent return to the wait sequence and repetition of the servicinq procedure. 9.3. Desiqn Principles of Multiprocessor Computer Systems (Co~aplexes) Special attention is now being devoted to developme.nt of multiprocessor computer complexes (MPVK) and in thfs case the processors of these systems have equal capabilities and they are MPVK assemblies. Conmunication betaeen them is accam- plished on the basis of standard interfaces. The main role in dsvelopment of multiprocessor computer vomplexes is to increase the productivity of systems by providing the capability of parallel fulfillment of independent tasks, inc asing the operatinq efficiency and improvinq the low dis- tribution in the system, providinq more economical servicing of outside assigr~ments and those with peak loads and achievinq a high effective utilization factor of re- sources to develop new types of configuratiion of the vomplex. The advantaqes of multiprocessor ooarputer complexes compared to multimachine c~- puter oon~ple~oes in speed are determined by the following factors: 58 FOR OF`F[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500060015-4 FOR OFF[CIAL USE ONLY - 1. Solution of fuadan~tal probleatoriented tasks such as autoaoation of larqe facility desiqn, workinq out different v~ersions of current aad future plans for development of the aouatry's economy and so, reduves to carxyisq out a larqe nwaber of calculatinq processes t13at utilize conmion data. The same sys- tems pr~grams (operatinq system, programa~ing automntion system), applied program packs and common data for tasks beinq solved simultaaeously can ~+e wsed for solv- inq these problems. All these data should be stored in the me~aory of ~~ach oonq~u- ter in a multimachine oomplex and there is r~o need to duplicate these data in the meaiory in a multiprocessor computer system. 2. Auxiliary devioes for synchronizinq these oomputers are sequired in a multimachine conQlex, each aamputer of r?hich oontrols its o~+a operatinq system, when solvinq several problems with conna~n data. 3. The efficiency of utilizing the capacity of the internal and external storaqe used by separate problems is oonsiderably enhanced in multiprocessor complexes. Most prablems require small mem~ry capacities durinq most time inter- vals and this capacity increases only durinq tQ~orary peak time seqments. Mem- ory capacity is selected accordinq to peak situations in single-processor system so as not to reduce the overall speed of probleursolvinq due to restrictions of the memory. Simultaneous solution oa different processors is possible in multiprocessor com- puter complexes when solvinq problen~s with smnll mem~ry capacities. If a sharp increase of inennry capacity is raquired during each time interval, the entire memory is set at the disposal of a sinqle problem. The main features of MPVR design include the following: the system includes one or sev~eral processors= the central metrory of the I~VK should be in oamaan usaqe and access to it from other processors of the system should be provided; the MPVK should have oonmioas access to all the input-output devices, includiag charniels ; the MPVK should have a~ified operatinq system that controls all hardware and software; interaction of the h~rdware and software components at all levels should be provided in MPVK: at the systems software lev~el, at the program level when solv- ing user problems (the capability of task redistribution), at the data exchanqe level, hardware interruption lev~el and so on. The latter feature af I~VK significantly distinquishes ~t from multimachine com- puter oomplexes, in which communication can be established only at the iaforma- tion level. Methods of connectinq diffe,.;nt functional units of the system to each other is of the most im~ortant significance for organization of a multiprocessor camputer 59 F'OR OF FICiAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 hUK VMtII�1A~.. UJG vl~Ll complex since tlie efficisncy of the MPVK is determined largely by tl3e deqree of parallellism or time combination of the operation of all devices of the system. Let us consider some types of orqanizinq multiprocessor computer systems. l. Systems with Conman or Time-Shared Bus This system is organized so that all functional units are connected to a single ~~,~n connectinq bus, which can have the width of one oomplete wvrd (byte) or one b~t. ~t?e less the width of the bus, the more comPlex is orqanization of system control. The bus that links nadules can be a single tw~o-waY (Figure 9.6) or they can be two one-way. Information is transmitted between modules thmugh the bus in the time-sharin4 m~de� The mem�rY is accessible to all the units of the system with this organization. IfaNOn Mo8 nb Moa ne 6 AoBa a~ (1) o~ ~~(2) l2 ~ d~Ou- ~ 1~ pa4eccop ( 3) Apoqeccop e~ � Fiqure 9.6. Orqaaization of System with Coannon or Time~Shared Bus ~y' 3. Processor 1. Input-output chaanel 2. Internal storaqe module Each information pack fed to the bus should oontain, besides the data subject to transmission, the address of the unit to which it should be directed. Conflicts among several packs fed to a single ~it cannot arise in this system since the bus contains only one pack at each moment of time, while all the remaininq information sources should wait until the bus is free, i.e., conflicts in the system are pre- vented automatically. Each unit oonnected to the bus should have equiPmei?t that recognizes the address in the pack. A system with a time-shared bus is very flexible and permits easy alteration of the number of functional units. Moreover, the system has low cost, low productiv- ity which is limited bien e ofeall avaiiableaunitss~AaioMPVKiwi~h c laar~onsor time- eous utilization effic cy shared bus is usually employed in low-power complexes. 2. Cross-Switchinq Systems ' A cross-switching N1PVK is Presentero esgoruor to7anY~inPu~outputPde ice.~Unlike tion of any memor7f module to any P 60 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500060015-4 switchinq with time-sharad bus, this m~tthod of switchinq is called the spnce- sharinq method. P'hysical oontact for the entire period ot iaformnticci traasmis- sion is established bet~re~ ariy tao ~ita aad the inform~ntion trausmiss3An speed increases in this case. Moreov!sr, several transmission paths can be established simultaneously. The switchinq n~atrix is caanpletely separated from the fuactic~al units of the sys- tem and can be desiqned by the modular principle, Mhich permits rather simple ex- pansion of it if the switchinq matrix has ac3equate capacitg?. But the switchinq matrix is in itself vesy oomplicated. Thus, for exam~le, a matrix for eiqht pm- cessors and 16 memory modules c~tains m~ra components than the processor. Oon- flicts durinq inquiries of the same mem~ry m~dule are resolved ia the s`+itching natrix. A cross-switched system has less flexibility oom~ared t,o the previous system. . - . hoay,re ~b~yiu ~ ~ ~ :~,,j~`' ~-x,, . . . ~ c~ . (1~ �p �i 4 . { . t . ; .,,~;ti�;'? ~acwec~ ~ yeao ' cop 0 . . �~z~~~~ , . , . ~ , � , . �x.~~~ . ~OO4er . yBBI . �.J coa ~ ~ ' ' ,r. h t � ~ t~r � . � ~ � ~ ~Z:~/ npoyec� deBP ~:r~,.. rnp ? ~ ..l . , , ;~'f ' . TovKU nept~tsoave~vv~~4) ~F~~ ~ Figure 9.7. Organization of Cross-S~aitched System Key: 1. Internal storaqe module 3. Input-output device 2. Processor 4. Switchinq p~ints An MPVK with cross-switchinq oonsists of function~nl units comparatively simple in structure, provides hiqh transmission speed and permits an increase of the sys- tem's efficiency by expandinq it throuqh increasinq the size of the switching matrix. Expandinq the syatem does not require alteration of the software. The main disadvantaqe of I~Vlc with cross-switchiaq is its oou~lexity and hiqh oost. 61 FOR OFF[C[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 3. lRultibua-Multi-input Systems Multibus-aulti-input ooaQut~r $Y~ms permit information to be transmitted over several lines simultaneously. The memory mcdules in this system should have several iaputs and should also ba equipped with ca~?tml circufts to rssolve conflicts in those cases whea tMO or m~re pz~cessors or iaput-output devices raquire access to the same memory modvle vrithin a ainqle cycle. Each processor and input-output coatr~l device in a mu].ti- procsssor com~uter complex of this type is ooetnected bo a separate faput of the central msm~ry m~dule by means of aa individual bus. The maximu~ number of m~dules cainected to the n~euwry unit is limited by the num- ber of inputs of the mennry module. An inPut expu?der or multiplexers can be used t~o increase the number of inputs.azid in this case the information transmis- sion speed should not e~oceed the permissible speed for a sinqle input. Canflicts frequeatly occur in systems with multi-input mem~ry, which are resolved by the afffliation principle, i.e., the priority of information transmission with simultaneous acvess is qiven to the ps+ocessor Which qains access to "its ovm" memory module (Fiqure 9.8). � ~poqeccop Ilpoqe:~ap , Moayns 0~ Noayne o~ MoByne Qll NoBya~ DI/ 2 yB9 ~3~ yBB Figure 9.8. Simpliffed Orqanizational Diagram of Multibus-Multi-input System Key: 1. Processor 3. Input-output device 2. Internal storaqe modula The width of the data transmission line is selected on the basis of convenience and econo~ny. If the mein unit of stored information is a word and the width of the data transmiasion line is smaller than a w~ord, then a special register for exparision and conwlution of words and also special ooatrol circuits that quaran- tee the correctness and continuity of transmitted inforniatioa are raquired. 62 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OF'FICIAL USE ONLY All memory units muat be conmected tfl each pmcessor of the system since each processor has its own memory unit in which special tables and control information are stored. This meth~od of connectinq the modules of the processor to the mem4ry m~dules reduces the reliability of the system since the capability of receiving information from the mem~ry module beloaqing to it and of continuing calculation~ made earlier by a processor that failed is lost with failure of any of the processors . The memory mcdule in all vomplexes of si~ailar confiquration should reoogrtize and process requests for access to specific mem~ry cells. The control circuits of the menbry module should resolve oonflicts during simultaneous access. The qreat advantage of these systems is that switching apparatus is not required for the processor. The complex can operate in the sinqle-pmcessor confiquration m~de, consisting of the same fuactional units without changing their design. The complex has hiqh information transmission speed. The disadvantaqes of these systems should include limitation on size and number of possible confiqurations determined by the limitation of the number aad types of available inputs in the mem~ry modules, large number of cables aad plugs and also the presence of expensive storaqe devices. 4. Computer Complexes~ with Mainline Psncessinq Increase of inforn~ation processing speed is achieved in the types of organization of c:omplexes considered above by increasinq the number of processors, i.e., by parallel fulfillment of operations. Organization of a computer systgm with main- line structure includes division of operations (instructions) inta several steps and in this case each step is perform~ed by a separate set of equipment. After the first step of the operation has been completed for the first pair of nwnbers, the unit that implements this step may beqin to fulfill the first step on the seoond pair of numbers and during this time the first pair of numbers is fed ' to ttae second unit, where the second step of the operation is impleme~ed. When the first pair of numbers is fed to the unit implementing the third step of the operation, the second pair of numbsrs is fed ta the second unit while the first unit beqins to pmcess the third pair of numbers and so on. An identical time to perform all steps of the operation must be selected with this organization of the complex. This time is assumed equal to the longest step. ~ The organization described above is suitable for a specific class of problems characterized by repetition of operations of the same time, for.example, arithmetic operations on a sequence of operands. Specially developed software and specific hardware are required to iaaplement it that ensure parallel completion of different elementary actions. The mainline method provides the ~reater effectiveness, the more complicated the operations to be perforn~ed are. The main advantaqes of multiprocessor computer cor~~lexes, due to which they find ev~r broader application, especially in compute: networks, should be noted: 63 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500460015-4 FOR OFFICIAL U5E UNLY high readiness and reliability= functional flexibilityi high efficiency an3 productivity. The readiness of the complex is usually understood as the degree of capability of the complex to perforcn its own functions at the moment when it is required for data processinq. Readiness time consists of the idle time and operatinq time of the complex. Therefore, this characteristic is closely related to the operating reliability of the complex. Recent investigations [22] showed that the reliability of a multiprocessor compu- ter complex is higher than that of a multimachine wmputer complex. This is relat- ed to the capability of access of all processors to f:he entire memory and the capacity of the multiprocessor computer complex to rapid reconfiquration. The flexibility of the system is not only the capability of the system to rear- range its own vonfiguration comparatively easily, but also to rearrange the func- tional properties of the system that ensu_e high readiness and reliability of the system with minimum conq~osition. This quality of the system permits all systems resources to be used by each user (task) and also provides access of several pro- grams to large data bases. Efficiency and productivity are determined in multiprocessor systems the same as = in single-processor computers. It should be noted that the utilization efficiency of multiprocessor complexes is especially high when solvinq large problen?s if a restriction is placed on tl~e time of solv:ing them. Solution of these problems may be impossible in a sinqle-processor computer. The main disadvantages of multiprocessor computer complexes should include the followinq: complexity of system program~aing; complexity of checking system programs and large expenditures on organization of checking; an increase of efficiency with an increase of the number of functional units is hindered by an increase in the cost of the complex; the capability of changing the confiquration of the complex and of increasinq the methods of communication is lir.~ited by the apparatus properties of functional units, namely the presence of time-sharing of a common bus, the number of inputs to the memory, specifications imposed by the interface and the presence of delays in transmission of signals determined by the lenqth of the cables between units. The E1'brus-1 multiprocessor computer complex having modular structure has been developed in our country. The following capabilities are provided in this multi- processor computer complex depending on its makeup: an increase of productivity 64 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FOR OFF7CIAL USE ONLY ~ Table 9.2. Number of Devices Sinqle~processor lp_g~~~X ~?pe of Device El-brus-1 E1'brus-1 Central grocessor 7. 10 Main storaqe 2 16 Main storaqe oommutator 1 4 _ Input-output prooessor 1 4 Data receiv~e-transmit processor 1 16 Synchronizer 1 1 Central enqineerinq vonsole 1 1 External storaqe Maqnetic drum and maqaetic disk storaqe control devices 1 4 Magnetic drum store 2 32 Magnetic disk store 4(8) 32(48) Magnetic tape storaqe contml devices 2 8 Maqnetic tape store 4(12) 32(64) input-output devices Punch card input 2(4) 8(16) Punch card output 2 4(16) Papertape input 2 4~8~ Papertape output 2 2 ~4~ Alphanumeric printer 2(6) 16(32) - Typewriter 2(4) 4(16) ~ Co~plex of four alphanumeric . displays (YeS-7906) 1~4~ 2~32~ Graphic inputroutput devices Graphic recorder (YeS-7051, 7052, 7053) ~2~ ~8~ Graph plotter maqnetic readinq device (YeS-7050) ~2~ ~8~ CRT alphanumeric and qraphic informatinn input-out~ut dsv3.ce (YeS-7064) ~l~ ~8~ Data preparation device Punch card da~a preparatfon device (YeS-9011, YeS-9010) 2(8) 16(48) Papertape data prepxra~ion device ~ (YeS-9024) 2 2~8~ Maqnetic tape data preparatf ~n device (YeS-9001) 2(4) 4(16) 65 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 l'va~ vr i� a~.a~+a.. . by increasinq the number of processes (1-10), an increase of internal storaqe ca- pacity (0.5-8 millinn characters and 4-32 modules), an increase of the number of input-output processor m~dules (1-4), essentially unlimited.expansion of ~he ex- ternal mem~ry capacity (magnetic drum, magnetic disk and magnetic tape modules) and an irncrease of the number of data receivinq-transmission modules (1-16). Sech module has built-in verification. The module is switched off if a~l~f~modules occurs. The qiven reliability of the complex is provided by redundancY of devices of the satne type (IQ~i) . The m~dular systea? provi@es the capability of adapting the comPuter complex to tlze m nt and al- class of problen~s being solved for mc~re efficient utilization of equiP~ae so permits a gradual increase and dev~elopsnent of the comPlex withont disturbing the operation of deviaes introduced earlier. A block diagran4 of the E1'brus-1 multiprocessor computer complex is shown in Fig- - ure 9.g and its makeup is presented in Table 9.2. The productivity of the E1-brus-1 single-processor system is 1.5 million opera- tions per second and that of the 10-processor complex is 12 million operations per second. The internal storage capacity is 576 Imytes and 4,608 Imytes� respectively. Nonaddress com~nand structure is adopted in the El'brus-1 comPuter complea~� This structure has qreat advantages. It provides : code compactness; simple and efficient translation of programs from high-level alqorithmic lenquaqes compared to the Unified Computer System; clynamic resource distribution; practically unlimited virtual memmorY; distribution of internal men?ory in segments of variable length; y hardware realization of standard alqorithms of the operating system. The central processor carries out dynamic distribution Af high-speed reqisters, distri.bution of internal storage in segments of variable length, orqanization of parallel processes, automatic calculation of index when processing multidimension- al files and workinq with fields of variable length and with numbers of different formats (32, 64 and 128 digits) during data processing. The operation of addition with fixed point is performed in the processor within 520 nanoseconds and that with floating point is performed within 780 nanoseoonds, multiplication of 32-digit nwnbers with fixed point requires 780 nanoseconds and logic operations and operations with fields are perfonaed within 520 nanoseconds. The compute~r COro essorsnt~Its produativ ty ssthree milli n perationseper second. of one of 1 p 66 FOR OFFIC(AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OFFICtAL U~~: UNLY algorithm. The operating system allocates one or several pages of the mathemat- ical memory (depending on the capacity of the file) for each data bit and the first w~ord of the file described by the descriptor is placed in the first word of the paqe. T'he capacity of one internal storage unit is 4K 36-digit words and access time is 1.2 microseconds. The internal storage is oonnected to the central processor and input-QUtput pmcessors throuqh the internal storaqe commutator. Each coaanutator organizes communication of four internal storaqe modules with the information user (the maximum number is 14). A fast channel unit and standard channel unit are used to exchange inforanation be- tween internal storage and external devices. The fast channel unit provides con- nection of up to 64 maqnetic drum and interchangeable maqnetic disk stores with sir~~~.taneous operation of four of th~n. There are a total of tw~o fast channels. The standard channel unit permits connection of 256 external devices (magnetic tape store and input-output devices) to it and 16 devices can operate simultaneously. The information exchange rate through the fast channel is up to four million bytes per second and that throuqh the standard channel is 1.3 million bytes per second. The E1'brus-1 multiprocessor computer complex is designed on the basis of inte- grated circuits and is related by its component base to third-generation computers. However, development of the principles imparted in this complex made it possible to work out and begin production of the EY'brus-2 multiprocessor computer complex with productivity of more than 100 million operations per second. It is designed on the basis of microcircuits with high degree of i.ntegration (large inteqrated circuits) and is therefore related to fourth-generation computers. The software designed for the E1'brus-1 multiprocessor computer complex can be _ fully utilized for the El'brus-2 complex as well. COPYRIGNT: Izdatel~stvo "Statistika", 1980 6521 CSQ: 8144/0540- B 68 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 unp (1) u~o N ~2~ K XC. M � ` ' ~ ~3~ ~ ~ on -I . . KM � ~41 ~ ~ ~ . ~5 ~a t6~ sw~ an Nn (8 � 90aneBNaa ~ an npxuyoae~ � ry~(9 -1 ( 1) 13) 13) 4) 15~i . I 10~ �K n~ ~a rn ~ L_-- ---____J Fiqure 9.9. Block Diaqram of E1'brus-1 Multiprocessor ~utsr pomPlex ~Y = l. Central p~ocessor 8. Remote terminals 2. domautator 9. inputroutput device 3. Internal storaqe 10. Alphanumeric printer 4. Input-out~ut processor 11. Punch card 5. Data receive-transmit processor 12. Papertape i,i. Magnetic drum and mnyaetic disk 13. Alphanwneric display device 14. Graphic display 7. Papertape 15. Graph plotter The high-speed memcry (SOP) of the complex is distributed by processors and is organized on the functiona~ principle. The hiqh-speed memory consists of five parts: fast vfrtual registers for storaqe of the top of the stack, local mem~ry with cdirect addressinq, associative memory for storage of the common dnta of two processors, associative memory for storaqe of frequently repeated procedures and buffcr mem~ry for ~pparatus advance pwapinq of data files. The internal storaqe has dynamic distribution and the mathen~atical paqes have a capacity of 512 words each. The inforniation bit of internal storaqe exchanqe with the processor has arbitrsry lenqth (with discreteness up to a word) and the length is determined by the descriptor that describes this data file on the basis of an 67 F'OR OFF[C[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/42/09: CIA-RDP82-40850R000500460015-4 ~ , , II~'ORMATIQ*1 CODING AND DATA PREPARATION Moscow EVM I VYCSISLITEL'NYYS SBTI in Russiau 1980 (siqned to press 20 May 80) PP 24-45. 72-104, 185-224, 303-3~4, 319-320 [Excerots from the book "Ooniputers and Ooaiputer Net~rks" by Vasiliy Nikolayevich Kriushin, Inna Nikolayevna Buravtsava, Nina Mikhaylovr~a Pushkina and Nina Griqor'- lnevna Chernyak, Izdatel'stv~o "Statistika", 18,000 copies, 328 pages] [Text] 2.1. Methods of ~ioodinq Econo~ic Iaformation for Machine Procesainq 2.1.1. Main Data on Machine Conversion of Informntion Any data that provides a ooncept of c~e or another aspect of the material ~rorld and of the processes occurrinq in it (natural phenc,mena, enents in social life, in techrroloqy and so ~n) is understood as informatio~n in the broad meaniaq of the~rord. Informa_3o~n is represented concretely by messaqes (in the form of speach, text, diqital data, images and so on). Thus, a messaqe is informntiou aa~bodied in n~ter- ial fonu. IKessages can be analoq and diqital. There are various types of information dependinq on the object of its use. One of the types of information is economic iafo~mation which is related to social produc- tion, distribution, exchanqe and coi?sumoption of mnterial gooda. it reflects phe- nomei?a and eventa occurrinq in the production and �oonomic activity of enterprises and orqanizations. _ Economic information perfornis the f~ction of planninq, requlation, moa?itoriag, accounting arid bookkeeping and has a n~n~er�of characteristic features such as: it is discrete since it reflects the activity of a facility through a system of natural and oost indicators rapressnted ia alphanumeric farm. it is characterized by a larqe meas character, c~mparativ~ely siniple mathemati- cal processing and the need for multiple groupinq aad saa~linqt 1 F~OR OFF[CUL [JSE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500460015-4 FOR OFF7CIAL USE ONLY initial information is vsuallp recorded i.n doc~nents aad is subject to rewrit- inq to machine carriers for oomputer entry. Evonomic informati~ oa f~ctions during meaagement is divided iato planniaq, ac- countinq, statistical at?d so ca. Discrete iaformation, including eco~omi.c information, is convenient bI? the fact that it caa be expressed by a set of elemeats of finite lenqth. Let us oousider the main cocicep~s and terms of machine display of iafor~ntion. The smallest inforaiation u~it is the bit (binary digit). Zhe elements which make up a diqital messaqe are called sysibols. A symbol is a qraph~cal sign that depicts the nwober, letter, serniee, matheantical aad other siqns. It is frequeatly called a syllable--a qroup of binary diqits used to rep- resent a symbol--as a machine uciit of information. An eiqht diqit syllable--a byte that has become the basic machine imit of information--is used to represent the symbol in the YeS L~VM [~ified computer system~. Zhe set of symbols for~s an al- phabet. The number of symbols in the alphabet is called the alphabet ~pacity. The binary alphabet havinq two symbols zero and one has ~eootoe widely used in coar puters. If number and letter values are assigned to information elemeats, the in- formation is called alphsn~aneric. Zhe aggreqate of symbols in an alphabet is a word. word lenqth is detern4ined by the number of symbols a~ntained in it. This is the num~ber of binary diqits in the binary system of calculatfon. A word is usually employed in machines to represent alphanumeric information. The lenqth of a machine word for a specific type of machine caa be either a oon- stant or variable value. For example, wrord lenqth is a specific value for first- and seoond-qeneration oom~uters. The lenqth of the alphanumeric Word is a variable value (1-256 bytes) in the unified coag~uter system. A machine w~rd ~ay be a binary word with floatiaq or fixed point, instruction or decimal nimd~ers. The agqreqate of sev~eral machine words cnmbined by a ~mified meaa is called a phrase or a recordinq. A group of phrases (recordings) sequentially arranqed on an external memory carrier, reoorded from an external meabry or read in it by a _ sinqle instruction, is called a imit. Several ~its containing recordinqs oom- bined by same feature into an infornation file is called a file. Input data must be transformed to the c3esired results when solvinq any problem. This action is called information conversion. In this case the input data and re- sults of calculations are expressed by different words but the reoorded data are expressed by the same alphabet. Informatiou can also be c~oaverted by translation fzvm ane alphabet to another. Thus, for example, using the syn~Uols 0 and 1, one can, as was shown earlier, convert nu~obers in decimal aotation to binary notation (a binary alphabet). Information conversioai in a computer is preceded by transfer of it from machine carriers or fmm different external devices. The results of calculations may in turu be transferred via input-output channels to external devices. Information is 70 FOR OFFICUL USB ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 entered intra a oomputer and ie ex~anqed betw~een the ooa~uter aad external devioes using specific codes. 2.1.2. Informatia~n ~ding To display informatioa in a oomputer, the alp~abet symbols of the ooaq~uter are ea- coded by a specific set of bi~?asy n~bers Mhich are arranqed in the diqit matrix of the ma~chine havinq a specific nviober of diqits. Zhus, codinq is oonv~ersion of information from one alphabet to another equivalent alphabet by usinq its symbols. T'he byte (syllable) ooasistfnq of eiqht bits (binary diqits) is used in the YeS L~VM as the basic minimal addressable vait of information. The oode of aa alphan~eric ~ synbol or two decimal numbers eacoded in binary-decimal notation or tr~o 16-diqit nwobers and any set of eiqht binary diqits can be located in a byte. Eiqht infori mation bits permit ocie to arrange 28 or 256 different symbols in binary notation (from syaibol 00000000 to 11111111) ia a byte. Othar nai.ts of iaforme~tion (half- w~ord, w~ord and double rrord) are multiples of the basic structural unit--the byte. The byte structure of data will be outlined in Chapter 3. Devices with various eacodinq systems that utilize oodes of different length are now used for data preparation. Individu~l systems are used to envode diqital and alphanwderic information. the most wic3espread are systems based on C~OSTs [State standardJ 10859-64, 13052-74, 19768- 74 and 19769-74. Three types of c~des--KOI-7, IOOI-8 aad DImI--are mainly used in m~dels of the YeS EVM. However, the presence of preliminary processinq uaiits permits information to be recoded from KOI-8 to DROI and vice versa in the u~ified computer s~rstem. KOI-7--a seven-bit code for iaformation exchange (GOST 13052-74)--is desigaed to display alphanumeric information at the inputs and outputs of data traasmission - equipment over commuaications lines and also to place infora~ation on papertapes. The table for encodinq symbols with seven-diqit oodes is sho~wr? in Fiqure 2.1. All the symbols of the alphanianeric vodes are subdivided into tWO classes--qraphic and control (special). ~raphic symbols include capital and lower case letters, nuaibers and symbols. Control (special) symbols include mainly oc~munications sym- bols which serve a control function when establishinq comnnmications and for de- termininq the formats of transmitted data. F+or example, PUS means empty, NZ means beqinninq of headinq, NT means beqinninq of text, !~T means end of text, KP means end of transmission and so on. The KOI-7 table consists of tw~o parts--the left part is called the first or Latin register and the riqht part is called the seoond or Russian reqister. A seven-bit binary code E7, S6, E5, E4, E3, E2 aad 81 oorresponds to each symbol of the table. The code of the symbol is determined by the positiun found at the intersection of the number of the colunn and line of the oode table. For ex,ample, the letter Ye in the Russian reqister position 14/5) corresponds to binary code 1100Z01 and in the Latin reqister (position 4/5) it corresponds to 1000101. The type of alpha- bet is indicated in front of the informntion by the symbols LAT or RUS (Vi~ and VYt~ in GOST 13052-~4). 71 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 evw vrra~.a~a. ~.s~ v~~a.� 1 ~ ' na~uNcKUU pe~c~p PyccKUU p%~utrp . 3~ 0 0 0 0 1 1 1 1 0 ~ 0 0 1 1 1 1 3~ 0 0 1 1 0 0 1 f 0 0 1 1 0 D 1 1 3s 0 J 0 101 01 0 1 0 1 O1 01 , ~ 3 s~ 3J 3: 3~ 0 f 2 3 4 5 6 7 8 9 10 11 f? 13 14 15 00 00 0 9CRPf '~RQQ H3C'J1 o A/l A . 0 0, 0 1 ~ y1 0 D 1 0 ? MT Cy2 ? Q R 6 T NT Gy? ? 6 p 6 P 0 0 f 1 3 KTCy3 ~ 9CSc s KTC9J ~ 3 uc UC 0 l 0 0 d K/I CTII lx a D r d t Kn trn T1 4 d m Q T 0 1 O 1 S NfT y SfVe u T NET y. S e f~ D 1 1 0 6~qp CNN b 6 f V V QN CMH i 6 p x~ p t 1 1 7 3B K6 / 7 6 W ar 3d X6 / 7 s/ I' B 1 0 0 0 8~t// RN ( 8 H X 1~ s BW RH 8 x� X 6 1 001 9 IT KH ) 91 Yf y fT KN ) 9 ueiMbl ~ ot0 ~OnC dM *~JZj zRC 3M u3 3 1 0 1 1 l1 BT RP2 K ~ k 8T AP2 + ; K ur K ll/ 1 1 0 0 l2 /14? PW + < L V C [ R~ P~ + < n 3 I! 3 1 1 0 1 19 BX P~ M] m BK Pl - = M W M i f 1 0 ty P3 >N A n- ye P9 � > M Y N y f l 1 1 ~S P3 0- 0 6 Q P3 / Y o- 0 6 Fiqure 2.1. Table for I~oodinq Syaiwls with Seven-Bit ~oaes for Informatioa Exchauge (KOI-7) lcey : 1. Latin reqister 3. S~ace 2. Russian register A seven-bit oode is the basic oode for data transmission equ~.pment and input-out- p~at devices of user stations, but it is inconvenient for inte~rnal processinq of inforaiation in machines. Because of this, the basic oodes for the YeS EVM are IoDI-8 (an eiqht-bit information exchaaqe code) and DIODI (biaary iaformation ex- chanqe and processinq code)--(~OST 19768-74. The advantaqe of these codes is the capability of enoodinq a larqe number of symbols (up to 256). 1�he KOI-8 is designed to exchanqe alphanumeric information with maqaetic carriers (maqnetic tapes, disks and cards). It is oonstructed on the basis of the seven- bit KoI-7 code. The seven smallest bits of an eight-bit oode correspond to the values of a sev~en-bit code. The valne of the larqest, eighth bit of the ICOI-8 de- tenaines to which reqister the code belaaiqs. If the eiqhth bit contains a zero, this means that the symbol belonqs tro the first (Latin) reqister. A value of one acoordinqly determines the affiliation of the symbol code to the second (Russian) reqister. The KOI-8 was developed with adherence to the recommendations of the international Is0 [expansion unknown] orqanization. The list of service and graphic symbols es- tablished by this standard provides the capability nr~t only of camputer and logic processing of information but also exchanqe of infoYmation over eommunications lines both aithin the count=y and over international communications liaes. 72 FpR OFF7C[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~OR OFF[QAL USE OM.Y ~ - ~ - ~ , ' 3oNOQan serpa8a (1) . , ~ oe 0 0 0 0 0 0 0 0 t 1 f 1�1 t 1 f ~ a, 0 0 0 0 1 t o o D 0~ ~ t~ E o~oo~~ao~roo>>oo>> ~ ~ Qs 0 1 0 1 0 1 0 1 0~ 0 1 D 1 0~ 0 0 0 o a~ a1 0 0, 00 Ol 0? OJ 04 OS 06 07 08 09 10 11 ~7 ~J ~i ~s . ' : 0 0 D 0 0 ~ P~ p 6 b n/0 A . ~ ~ 0 0 0 f 1 NJ 1! 1 R 0 q q A A ' � 0 0 1 0 ? M y? B R ~ T PA 6 6 ~ D 0 1 1 3 KT A~ 3 C S c S 9 u Q 0 1 D 0 ~r X/1 lx 4 B T d t 6X d m q oo~o~ s sEUeuNC y Qor ~o F a ~sFV v Hen ~~+rw �'o r > > ~se~r~�~~cw w r er ~ Q 1 0 0 0 8 RH ( 8 M X fi ,t ~ s , b - , 0 1 0 0 f 9 fT KH ) 9 I Y i u e? N bl � ~ 1 ~,1 D 10 ~C JM J Z j, 2 yP u~ li J 0 i i 1~d K[ k K m ~u l 0 0 lP A~ L~ ~ ~ n a/1 3 . 1 1 D 1 I3 BN P~ M J, rn M~K l!( ~�~1 t t 0 ~f 3�> N A n- N v V 1 1 f t ~S N P3 0- 0 36 i 1 a 6 Figure 2.2. Qode Table of Eiqht-Bit Informatiaai $xchanqe and Processinq Oode (lmi-8) Key: 1. Zone tetrad 3. Space 2. Dfgital tetrad The DROI is desiqned to process infozmation fn maichines. It is oonveuient for placinq infoxmation on puach cards. The IODi-8 and DKOI oode tables are stawn in Fiqures 2.2 and 2.3, respectively. As can be seen from the tables, there is no reqister division in them. The bits of the code combination are deaoted by the letters ag-al in the ROI-8 and by the numbers 0-7 in the DI~DI. The letters and numbers indicate the ordinal ntmibers of bits in the oode combination and oorrespond to the follv~a~ing weiqhts: B~ ~ KON-8 1 ~ Q~ ~ a~ ~ Q~ ~ Q~ ~ a~ ~ as ~ aa ~ at 6Mm s j(KON (2) 0 ~ t ~ 2 ~ 3 ~ 4 ~ 5 ~ 6 ~ 7 Beca ' ~ 2' ~ 2~ ~ 2~ ~ Z~ ~ 2' ~ 21 ~ r Key: l. Bits in 100I-8 3. Weights 2. Bits in DKOI 73 F~OR OFFZC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 evw ve~a~,u~a. v.,c. .i,~a.a 3oNO0~a � rerpu o oooooAooo>>i>>>>> 1 D 0 0 0 1 1 1 1 0 0 0 0 1 1 I 1 2 0 0 I ~ 0'0 1 l 0 0 1 1 0 ~ 1 t 3 0 T D 1 0 1 0 1 0 1 0 f 0 1 0 1 0 1? 3 4 S 6 7 7 3 4 S 6 7 8 9 R B C D f f 0000 o v~ ,s a- uvA6 .o 0 L~ D f 1 N3 1 7 ~ Q -W q ~ ~ 0 0 1 0 ? Nl PI! Q 1F S J a' k S 2 0 0/ 1 3 t C C t m C G T 3 0/ 0 0 i 6M d rn rt 3 D N U 4 2 0~ 0 I S fT l/C OC UCy e n v w E N V S ~o U 1 f 0 6 II d X6 8~ q o Lr v F 0 W 6 a 0 I 1 1 7 36 MI! p= a 6 P X 7 ~ ~ ood 8 AN t~ 6b roNor 8 , 1 0 0 1 8 MN ~ 1 r z ! R Z 9 e 1 0 1 0 A 9P 8 K 6 3 �Q ~ 0 1 f B dT f ~ tt , m n u N y u/ ~'f ~ 0 0 C va~ n< ~Y.(a~Mmq Nnxt3~ y 1 1 0 1 D~If P~ l)- i E N A 1 f 1 0 f PJ l1 ~t ~ p b 4 1 I 1 1 F ~3 3Q dN ! n?.+ u n 6 I' b~ 36 Fiqure 2.3. Qode Table for Binary Information Exchanqe and Processinq Oode (DROI) Key: l. Zone tetrad 3� - 2. Diqital tetrad The eacoding byte is divided into tetrads. The four most significant bits of the code combination depicted above Che ordinal number of the table volu~m (ag-a5 in KOI-8 and 0-3 in D10DI) are the zone tetrad while the four least significant bits lxated alonqside the ordinal nwaber of the raw (a4-al and 4-7) are the diqital tetrad. The first tw~ most sigr~ificant bits of the DKOI in the zone tetrad deter- ~ mine one of the zones: 00 zone of service symbols= O1 zo~ne of qraphical symbolss 10 zone of lawer case Latin and Russian letterst 11 zone of capital Latin and Russian letters and numbers. The next tw~o most significant bits, alonq with the first, determine the colvu~ in which the symbol is located. The bits of the least significant diqits of the byte finally determine the address. Any symbol of 256 code combinatiezis is determined by the code position (a fraction whose nwnerator is the column number and whose denominator is the row number) and also by a code combination--the code. For ex- ample, the code of the capital letter E is determiaed: by a fraction--04/5 (KOI-8) and C/5 (DKOI); ' ?4 FOR OFE'ICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/49: CIA-RDP82-00850R040500060015-4 by tha cod~s coaibinatioa--O1000101 and 11000101. The code for the syml~ol KT (and of text) ~+ill b~ detarmined as 00000011 (for 10~I-8 and D1oDI) by the sams rules. The machine w~ord of the xOI-8 will be encoded by fiv~e bytes and Mrittea in th~ machine mea~ozy ia the follo~ving fozms ~ K O N - 8' KON-8 O1001011 � OIOOtIII 11101001 OlOlllll OOIII000 AKON - I1010010 11010110 IIOOl011 01100000 I1111000 2.2. Information Carriers An information carrfer is understood in the qeneral sense as the phyaical aediu~n ~rhich is used to revord and store iaformation. The basic economic information carriers are the primary c3ocuments in Which all the ec.~oaomic operations completed by an entarprise or organization are reflected. They include different requirements, overhead, orders, iaterchaaqeable certific~*_a~s and .so on. Ho~vever, primary docwoents are m~stly suitable for use by mau but ara not suitable for direct co~uter entzy of data from them. Therefore, the informat~on from them is first placed on intermediate machine carriers in oodes suitable for oomputer processinq usinq data preparation devices. Pu~ch cards, papertapes, maqnetic tapes, cards, drums aund disks are used as the machin~ carriers of informa- tion. Besides these carrfers suitable ~uly for use in oomputers, carriers that combine primary docwnents and machine carriers, which are suitable for simultanecus use by man and ccmputers, find application. They include machine-readable docu- ments--dual-cards, standardized forms the infnrmation fro~m Which is read directly by means of automatic reading machines, punch cards with graphical marks and so on. Let us consider the basic machine carriers and methods of encodin~ infoYmation on them. ~o� 3 ~ t~4 2 , ~1) j 12no~uquw J,t .r'~7., fl t~~ ~ Nttu~~ r~tusxr U~qr~~sKr~MYpMt~M7171~a~� _ ~1~~11U~1~~11111 1 1 ~~~1~11~111~ N11111~~~~I~~~U~~t11111~~~1~1~~~1111~~1~~ 0-"- ~i~ttilitili~Mti~ t ~ ~i~~~l~ll~l11~ ~I~t~ttAtltillf*111ftlli1w11~IIi11i111f1111111M ltTIJ2~I2il27ttt t! itittstt!l221t t!!tltt2i2t22!lZt22!=lttZ!lttttil7~lt2J~2lt ~ - ti 7)11~71~1177~7)) )7 )1~173~1777777 I7I777I�>I7777�77717l7791~7777~1~7~7~7f17 - � ~ I~~~~~/~11~~1~1~ 1111~111111111 IIII/It1/11/11111111111/{/IIIIt//11111~11~~ 4 - ~ - flSSflSlfSfSSSSSSSf SSSit77lS3377S l~iS~fSSlSS37~!!l777l7lSSl7~SSl3l73l1777777 S - � - - iti~iii{fifiillf~ii iiliii{{iiiili liifi{fiiifliitiiiiif{lili{iii1i1111f{{{{{i 6 - ~r~�~,��n�>na �?mm�� >�r~~mn~nnnmm��m~t~��n�r n~uuuuu~~t~u~~uuuuu~u ~~ui~~u~~u~~~~~~~uu~tuu~in~~~uu~~ 8-+- 1 2 ~ 1 ~ t~ It 1~ N N 7t f~ ri f~ tl 71 f1 f 11 I Y t~ L 9 Y 1~ 1~ M P M M N 7~ 7t T~ 7~ M !!!fl11!!!!f!1l111!! 1fHf1111l1! 1f1ffllll!lfff!!!lf~1lflfl1f11!!!!f!f!!f!! 9 ! ~7.2! / 7 J Figure 2.4. 80-Column Punch CarcY (P1G80) Key: 1. Position 75 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 rvw vr v~+~. s The punch card is a recta~qulax paper card measurinq 187.32 X 82.5 mm (GOST 6198- 75), manufactured from Pvnch oard Paper (0.18 ~ thick with tolerance of + 0.01 m~n. The upper left hand oorner of the puach card is cut at an aa91e of 60� for conven' ience in stackinq it into files. The remainit?9 corners of the punch card ~a be straiqht or rouaded for qreater wear resistaace. R~o types of punch cards are naw emPloyed--45-column (PR45) and 80-columai (p1cB0)- Eiqhty-col~n puach cards havinq hiqher recordinq c3ensity are mainly used in oom- puters (Figure 2.4). The numerical matrix inQrinted on the fmnt side of the punch card separates it into v~ertical x+o~ws--ool~s 1 and horizont~l rows--positions (lines) 12-9. Be- sides the 10 pcsitions denoted by nwabers, there are additioaial llth and 12th positions not denoted by nu~abers in supernumerical field 2. The ~o~t~ on tended for imprinting on the puach card (input d~?ta ana Programs~ forms in octal or decimal oountinq systems. It is then transferred to the Punch cards in the forn? of perforators 3 located at the intersection of a positi,on and columc~ by using data preparati~n devices. In this case the perforation corresPonds to oode 1 and the absence of a perforation corresponds to code The Perforations may be numbers, letters and special symbols whi~~. are automntically read and trans- fe rred to the co~nputer memory . Acoordinq to GOST 619~-75, 80-coltnan punch cards are manufactured in the following types: pK80--with oomplete numerical qrid= PK80N--with inaxnplete numerical qridf PG80/27--With zone for graph mnrks (oae-sid~ed) s PG80/54--with zone for graph marks (two-sided). C CT ymlu YIUYtA?0 310p ~ ~ ~ R6 AlpEif3 Nu MlIMN00~ f~t~~sc~e~~�~ ~ ~ t t M ~ 1 1~ 1 1 1/ 111 ~ 1 1 1 1/ 1 1~~ 1~1 1 1 ~ 1 1 1 1 1 1 1 1 ~ 1 � 1 1 1 1 1 1 1 1 1 I11 t~ 1 1 1 1 1 1 1 t 1 1 1 1 111 1 1 1 1 1 1 1 f 1 1 1 1 1 t2t 2 t 2,! 1 2 t lili2 2 2 t 2 i 2 I 2 I t t 2 i 22Z2 t 2 2 2 t t i I t=__ 7~l~1 ~ 7 7 7 I)77711 ] l>> ) 7 7 ) 1 3 ] 7 7777~77 3)~ 7~ 7~] 7 ~I1~111 ~ 1~~1~11~11 1 1 l 1 1 1 t~ 1~IIf11~~~ l~~~! 1 1 1 ~!!!lSS777 S S S S 3SSSSSSSSS! S S Z S S 3 S S i S SSlSSSSSSSS S 7 7 S i 5 S J f fiiitiiif~i i i i fiiflfiiiifii f i i f f f i i f ififfii~if{i{ f f i f i i i f 7/)f7I7)f7~17 7 1 777717~77ff777 7 7 7) 7) 7 f 1777771777I717 7~ 7 7 7 7 7 ~i~ u~u~i~~~e~ ~ ~~~~t~~~~~u~e~~~ ~ ~ � ~ ~ ~ ~ ~~~~~~i~~.~~~~~~i ~ ~ ~ ~ ~t ~ ~��~eun ~�nnnea~++~s ra W�r1~lM1~lM cturw a�s �N , 11!!!!f!!lf~1!!!1 !lffl!!!llff!!!!!fl~11f!!!ffl1~lflflflff!!!llfflf!! !fflf!!!!! . Figure 2. 3. Encod3ng of Alphamnaeric ~Infonaation in Decimal oounting S~istea~ According to GOST 10859-64 76 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 N~obers are printed only in positions 0, 3, 6 aad 9 in punch cards with incomplete numerical qrid, which permits auxiliary inscriptions on the frr~nnt side of the pnnch card. The arraaqemeat a~nd dimensions of the perforations in puach cards should oorrespoad to GOST 8912-76. Rectangular holes are puached with base of 1.4 ma~ and height of 3.2 m~ ia an 80- oolumz~ punch card. The distance between the centers of the perforations in two adjacent oolwons (the spacing betw~reen columns) is 2.21 mm. The distance between the ce~ters of two adjacet?t perforations in a single coluta~ (spacinq between posi- tions) comprises 6.35 a~a. The maximum number of perforaticu?s which can be placed on aa SO-oolua~ pu~ch card is 960 (12 X 80). The information on punch cards can be arranged by vol~s (in Minsk and YeS com- puters) or by position on liaes (in BESM and Qral computers). Acoordinq to GOST 10859-64, infonaation can be inserted in decimal or binary n~ber systems. One symbol usually occupies one columa~ with this arrangeanent. The code of GOST 10859- 64 permits one to encode no m~re than 128 different symbols. It is made up so that any symbol in the colunm is depicted by an odd number of perforations (1, 3, 5, 7), which is necessary to check the correctness of data entry into the computer. Z'he zero position of the punch card serves as a supplement to oddaess. The punch cards are inserted in the readinq device by the narrow side with colw~ arrangement of infozmation. Representation of information in the decimal nu~ober system is used extensively when processing economic information. FSncodinq of alphanumeric information on an 80-colwa~? punch card in the decimal number system acoording to GOST 10859-64 is shown in Fiqure 2.5. If the infonaation is arranqed by position (by line), one symbol xcupies from three to eight coluans of a single position (line) dependinq on the type of in- fonaation. The punch cards are inserted in ~he.readinq device by the wide side. Tlie 12-position I~K-12 punch card code with col~nn arrangement of symbols is used in the YeS EVM to enter inforniation an the punch card. The I~K-12 code table is presented in Fiqure 2.6. The table is divided into two parts--a lef.t part (col- umns 0-7) and riqht part (columns 8-15). The code combinations that indicate the . nu~nbers of the positions in the punch card in which perforations should be made are presented in the e~reme left hand and riqht hand vol~anns and also in the top row of the table. Letters of the Russian alphabet, similar to Latin letters, have the code of Latin letters. Instead of the symbol (code position 1/15), the symbol is permit- ted. The capital letter b is permitted in code position 7/7. Any symbol of 256 code combinations is determined by the code position (the fraction whose nwnerator is the col~nn number and whose der~ominator is the row nwaber (and also by the code combination--the code. For example, the code of capital letter W is determined: by fraction 15/llt by code combination--12-11-0-9-8-3. 77 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/02/09: CIA-RDP82-00850R000500460015-4 hUK UMNII:IAL UJr, V'VLY 7 17 17 t: 17 1 17 1 t ~ tt l? II !t o 0 � p 0 0 0 0 0 N~ 0 1 2 J ~ S 6 7 A 9 10 if f7 13 1S . . o a- o ( i . u a a-, ~ r p J i r a j-.~ u3 cir? po~ An 1.3i ~ t e rr s t e~ s~~r~ cyi ~oa twH s-2 J J C L T ~ 3 C l i w RJ G5J ,QOJ /1~9 0'J ~ ~ D M ~ 4 d m u a,C1F Q79 ~104 /f1a 9-4 S S� M Y S t n u W!: i~05 /!C ,~71 9-S 6 6~ 0 N 6/ o r v Q06 BIU xb Q77 b !-6 ~ ~ 6 v x ~ o p.c 3 35 Q� ~v~ Kn 9-7 8 d N Q i 8 h q y~C ~I3 ~N QOd J/7; 6 ~ 9-a 9 9 1 ti 3 � r i f7'~ xN ~'~9 Ac'~ 09C Fvt ~QD Q:f 9-d-t 9-t ~0 [ 7 ~ : a x 6 4fi ~iB ,R~O :;:'6 3 9-8-2 a-3 11 . ~,,N p I~ BT ,Q~S Qt1 G2~ M Y~!f/ 9-d-3 d-~ /7 y. w~ ~ M 7 l! f~a P~ A~2 CTI? N /1 J1l 3 l-d-4 e-S 13 ( l t w BK P/ X7N Nfl A G~ 9-d-S 8-6 x Q! ~yC R3 ,QR ~730 R 6 4 9�I-6 a-7 ~s ! ' u n! r nar v3 3a JM 6~ 36 9-B�7 " Fiqure 2.6. Code Table of 12-Pos~tion I~K-12 Code Key : l. Space f~~S79-uKHWS~binpna~t�yNT6++~~t[~(:CKNCZ~YRRLV ~ fY 680N9E~ X 811OQ3AG b/~!* /)JUE6jHF POD K ~ ~ ~ ~ ~ R ~ 1! ~ j II 2 2 2 2 2 � 7~ l 7~ ] 7 7] 7 u� ~ ~ ~ u ~ ~ ~ ~sss ss s ss ss ~ s s sss~ s ss fiiii Ii ii f{ ff i f 6 iiifii i ~fi 71777) 1) 71 71 7 77 7 1 7 177777 ~ 77 ~77! , ~~~~~ul ee e~ ~i ~ ~ ~ ~ue~~~ ~ ~tt~ ~ t ~ ~ ,i u u u u a x ~ o u o II!1 f{ 7~ M Q M M n n rt 7~ r ~ 7!!!!f!!f f!!! ff f ! !f ! f!7!!!f!f!!!!!!!!7!!!!!! !!!f!!! . Figure 2.7. Representation of Alphanumeric Information in I~R-12 Oode Representation of information in KPK-12 code is shown in Figure 2.7. The exten- sive use of punch cards as information carriers is explained by their advantaqes: high strength and long storage time; - convenience of making corrections to errors by punching new cards; 78 FOR OF'FIC7AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500060015-4 FOR OFFICIAI. USE ONL ~ - the capabi.lity of creating a large capacity of penoanent card filess convenience of qrouping by features and retxieval of the required infoYmation; The disadvantaqe of peuich cards is the limited recordinq density l'80 decimal diq- its) and the speed of reading informntion (300 punch cards per minute when fed by the narro~.~ side) . Papertape is a narro~ tape 0.1-0.15 nm~ thick manufactured from hiqh-quality sul- fate or waxed paper. Tapes of increased strenqth manufactured frcm plastics are also used. Znformation is entered on the pnpertape by punchitiq round holes arranged alang the code tracks (positions). Five-, six-, seven- and eiqht-track tapes are used. The perforations in the transverse direction form rows. A single symbol corresponding to the code of a number letter or siqn is written in each row. The maximtmm num- ber of different symbols which can be placed on a papertape depends on the nunber of coding tracks which in turm detezmine the type of code used. For example, up to 32 symbols (25) can be written on a five-track tape and up to 256 symhols (28) can be wx~itten on an ei3ht-'crack tape. Three types of code--the: second international code MTR-2 intended for data trans- mission over comnunicatians lines, GOST 10859-64 code used mainly to enter infor- mati~n in a computer and KG~-? code (GOST 13052-74), which is unified for entry of information into the ao~uter and transmission of it over comnunications lines. Kvdodn~e c'opoxKU ~ ~ ^Q ~1 Q, ~ w ~4'�=' N N ~ w ~ O Q ~ ~ ~ tu - ~ ~ ~ ID +i +~1 p ~-1 U ao ~ tVtl ~a/! v~1 ~ C C 4 i~i ~ ( fAN +i O O ~ v$~ ~ m u~ ~ ^ ~ ~ ~'C 'O ~ ~ � C ~t~ ~ 1� - a u~,~ O~ b~ C .~~1 O aD ~ C~ ~ Yj ~ i~ W~ e~Oi 3 O ~ I~~a ~ N ~ ~ ~ fA W W 8~y W W~~A a ~~q a~ e~ w$ a~ ~~EEp+~Q~~EES~ i7 tA ~ W a.., . � . . � � . . . . . LL V a Cr ~ ~ ~-1 N e+f tT IA t0 1~ m 01 O.-1 E.Ip, V ~ ~ ~ ' N 'y� ~ ~V ,N rl ri .-1 N N ? ti V U . w y '~~c ~ a a) V N ~ ~ ~ N u'! ~ u~~.u - ~ ~ a 'n b ^ ~ O ~ - N c, V 0 ~,5~ N ly ~ ~ N $ ^ W ~ ~ W l~ CV ~ ~ ~ 'r ~ YI ' , V ~ ~ ~ ~ V M'~, V ~V CV ~ ~ ' N ~q ~ ~f! A1 W r~U{) ~ U~ A ~V ~ ~ ~ , ~ ~ ~ r? ~ o a ~r aa..a.~ a ~ a v, ~n rn a w.-~ m N ~ ~ : ~ ~ 9 I ~ a ~ ~ c~'i ~ ~ a� ~ ~ v Y Y ~j = O 71~ J,) tiv vvvvv v S R ~ s RN b M 4 s ~ � ~ o cv b~~n ~ n v~ o e~ ~ ~ ~.Y s * ae a a ~I cV r G Y u ~ _ ~ w 00 i~100000r10 +~1 ~ 1A 1f1 b IA tf1 IA 1A tf1 IA ~ a e`~a ~ a Y =~a m u> v~ m~ v~i ~~i ci~ v~ ~ ~ ~ a � ~ o ~1 aQ m ~ Y s r1 C C S~' v 0 f7 ~ f~ f~ 'i~ ~ ~ ~'v ~y F f~+ S e*i S S C S ~ S 7 S ~ � � � � � � � � � e v;.,� o 0 0 ~O.~o 4~ ~ _ M.-I N rf ~t u1 ~C 1~ ~ G? O ~ r U I U C ( j~ = q ~ Y ~-1 ~ 155 FOR OE'1~7CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007102/09: CIA-RDP82-00850R000500060015-4 l'VK VMMII.~AL VJG UhLY for access to the tracks on each storaqe device. Interchanqeable disks of the YeS conQuters have reoordinq density of 30 bits/am oa~ the ou~side track and 44 bits/mm on the inside track. A ftmctional diaqram of a maqnetic disk storaqe device is shown in Figure 8.13. Magnetic disks 1 rotate on a drive shaft 6 at a speed of 1,000-3,000 rpm. sprin9 levers 2 with magnetic heads 3 are lxated between the disks. The levers are riq- idly attached to a carriaqe 4, which can move along a guide shaft 5. ~rie maqnetic heads m~ve aloaq the radius of the disks within their ciwn zone 7. ~ ~ ~ 3 6 . ~ s � . Figure 8.13. Fu~ctional Diagram of Maqnotic Disk Storaqe Device A diaqram of the interaction of the magnetic heads and the YeS-5053 interchanqeable magnetic disk pack is shcwn in Fiqure 8.14. This pack consists of six disks whose outer diameter is equal to 336.4 m~n. The surface of the disks is coated with fer- rolaoquer 4-5 microns thick. Working disks have a thickness of 1.27 mm. The distance between the disk surfaces is lO.lE mm. The YeS-5053 pack pes~ni.ts recording on 200 main and three reserve tracks on each surface of the disk. Tracks with addresses 200, 201 and 202 are used as reserve tracks. The track with address 073 is reqarded as the verifica- tion track. A total of 3,625 bytes can be recorded on a sinqle track. Ten tracks arrange3 one under the other on 10 w~orking surfaces of the disks fona a cylinder. The capacity of a single cylinder is 36,250 bytes while the capacity of the total pack is 7.25 l~ytes. The disk pack rotates in the storaqe device at a speed of 2,400 rpm. The storaqe device has 10 universal maqnetic read/write heads (one head each on each working surface) and 10 erase heads arranged in a sinqle housing for writ- inq and readinq infonnation. Only one of 10 heac~s operates simultaneously when ~ritiaq or reading information. They write (read) information on the same track digit by diqit. The heads are numbered from bottrom to top from 0 to 9. Z'he address of each track in the pack is determined by the cylinder address aad the nwaber of the head. The time recorded to retriev~e and transmit data consists of the cylinder retrieval time, head gelection time, carrier rotation delay time and data transmission time. The total time of the indicated components is called access time. 156 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 S' /~uY~ AaD � 211Dt ~g~ ~ , . ~ E~~ ~~Tka : j ~ t7 ~R~ ~ H /~cs . aovro~d4ru _ - QucKV (5~ Fiqure 8.14. Diaqram of I~teraction o! l~laqaetic Heads and YeS-5053 Maq:?etic . Disk Pack ~Y= 1. Gylf aders 4. !~lag~etic heads 2. Carriaqe 5. Disks 3. Track The erase head is arranged behind the read/write head in the direction of sntation of the disk. Durinq data recording, this haad demaqnetizes the intervals betMeeri adjacent tracks, which reduces the mutual interferenva from these tracks duriaq subsequent readinq of information. The YeS-5050, YeS-5052, YeS-5055, YeS-5056 and YeS-5058 msgrsetic disk stores oper- ate with interchangeable maqnetic diak packs of type YeS-5053. These stores have ideritical main specifications, namely: capacity of 7.25 I+~ytes aad data transmis- sion speed during writinq and reaa1inq of 156 lmytes/s. They are distiaquished mainly by the av~eraqe sccass tims (see Table 8.5). The YeS-5061 Nl~ID, `+hich aperates with YeS-5261 iaterchanqeable disk pack, has a siqnificant diffarer?cs. The YeS-5261 pack differs from the YeS-5053 pack by the fact that it oontaias 11 rather 6 disks (20 v~rkinq surfaces) vrhile each cyliadsr contains 20 tracks. Raoording dsnaity is 60 bits/m~n on the outside track and 90 bits/mia on the inside track, which pesmits recordinq of 7,250 bytes o~n a siaqle track. The capacity of the pack is 29.17 t+mytes an8 data transmieaian apeed is 312 lmytes/s. The mcst productive interchanqeable maqnetic disk store is the YeS-5066, in Mhich a pack of 12 YeS-5266 disks is used. I~ecordinq is accomplished on 19 surfaces of the disks. There are 404 main and 7 reserve tracks on each surface. A total of 13,030 bytes is reoorded on each track with rewrdinq 8ensity of 160 bits/na~. The pack capacity oanprises 100 Irbytes . Zl~e data transmission speed is 806 l~ytes/s . - Permaneat maqnetic disk stores include the YeS-5051 (USSR) and YeS-5060 (Hunqarian Peoples Republic). 157 FOR OFF'ICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 m ~ m a ~ ~ ~ s a ~ N ~ ~ ~ o .~.~1~ ~p ~ ~ = V ~ yV,l ~ R3 ~ ~1 ~ C~ ~ ` r1 V ~ ~ R7 ~ a ~ ~ dl ~ m +~~1 ~ V O o~m c o N_ I g 5~ ( p, ~ o-~-~ :y- _ ~ .~~m> tA v ~ U ~ Es � S~ ~ " ~ ' R a $ = 'D �-�i o m m m ~ t~i ~ "~~T T ~ ~ ~ a ~ a ~ a~o m ~ w a o~~~~~mcmi~a ~n ~I ~ o o ~ ~ ~oooi0b~ ay. u S c~ I~: ~ _'n U o.c s~ s+ ~+~i o y v ~ ~ w i'a m amm s+~ao o ~a ~ O~ ~ b~~~ > b O O ~ m ~ ~ ~ a $ ~ ~ ~+oa~?-~~c~zzz~eou~ V V � � � � . � � � � � � � rl Nn1d'Il1tGl~~C?O r-1N ~A ~ U = ~ ~ N ~ .-1 rl r/ r-1 ~-1 .-1 rl rl N N N X ~ a a N ~ ~ � ~i w ,~r~ ~ ~a c~ ~ p tVU U ~ cv ~ ~ U ~ ~ ~ `~"a a c. O ~ N tD ~ ~n ' ~ _ ^ w ~ ~ .=a st'~ ~ ~ in Ol ~ v ~ N ~j N N N] ~ V ~ ~ ~ T ao w . ~ ~ a U ~ ~ n 5~ o N ~ a a a , ~ ^ ~i - I o~ ~ ~ m I ~ T I N W o~ m ~ ~ ~ ~ ~ b I ~ ~ ~ ~ ` ~n a a a a a~a a ~ I= I I~.~. ~ I.. I " p vi vi cZ v~ rn ~ ~ A ~lf tA CI ~ ~ ai ~ O~ ~ N ~.~1 D~ W ~0 U' ~ U W d x U I;~ p K q~ ~I .{J v v v y.~ v v v v.~ v v ~ q = Y X a U O�~ N r0 i[1 ~O CD rl 10 O U s ~.~I 10 Uf u1 ~c1 b~ u1 tA Il1 t0 t7+ t0 +~1 ~ ,v = _ 3 ~ = s W 000~0000.-10~-1 i ~ ~ - �J ~1 ~ lf1 ~f1 IA 'y 1A tA t[1 tA u1 A m �J 1O ~O m V 1 1 1 fA 1 I 1 1 W 1:J ~ ~v C ~y s A UI ~A fA [A (!I N~A (A W . ! rl rl aYi 4 - c ~ y dl 41 41 W 41 41 al O W m ~ ~ v v Y ~ ~ a, NA~~~ 0~~~~ O~ ~ ~ ~ _ m _ ~�o a�~ ~t `t ~t m ~ r` ~~C n a C y ~ H � � � � � � � � � � Y 4! ~ ' t U S ~ C! r.'. pS P'~ N ~ ~ ~ ~ ~ ~ ~ ~ A i r" p < rm- � ~ 9 ~ o. ~ ^ Z ~ Y ~ u u N 61 H s u w s T s ~ ~ ~ I v~% i s o 0 o a ~ o F- 47 :~L X :+C V U X i58 F'OR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ~ ~ The YeS-5051 store has a pern~anent msqnetic disk pack dividad inbo t~ro uaits of 18 disks each. Storaqe capacity is 125 lmytes, the nun6er of Mo~kinq heads is 192 and the averaqe access time is 250 ms. The Yes-5060 store has a sinqle maqnatic disk vrith fi~oed heads. It is desiqned tc work with the Yes-1010 computer. The outer diameter of the disk is 31'7 am anc7 thickness is 9 ain. Disk rotation speed is 3,600 rpm. A total of 17 magnetic head units, in each of which there are 16 `o~orking and caie reserve head, are arranged ou both sides of the disk in the store. Rev~rdinq density is 40 bits/mm. Disk ca- pacity is S00 Imytes and information exchanqe speed is 150 A~ytes/s. The specifications of maqaetic disk stores used in models of the YeS computers are presented in Table 8.5. An identical track format (Fiqure 8.15), which is deterae~ned by the control device, is used in irodels of the YeS computers. There are sections on the disk txacks for recording operating and control information. Ccntrol information is used to organ- ize and v~erify direct access and contains track and revording addresses, lenqth of recordinq and intervals between zcaies. ~?'11~~ ~'1 . . 3G0(ICb~, ' ~ r~~e ~l~ ~3~onHCArEhe. ~6~ ~ NMBlNCNN(! ~2~ ,Q~vG~wK ^ Janucs R~ � ' 3anucb R� MapKep ~ ~ ~ pp~ M~ C,ret nt ~ii Ay cvl~ Il~ Mae n~ y ~ nw ~4~ (5) , ~ , ~ , . ~ :;f,:, . ~ . m ~ r 4K ~ q~ ~ s AK . a0 uK � quK~u� . ~ ' . MlCKlH/ ~ KON7QQ~ l~(IK/!!I. /luN~l. VMA'IIfI ~1~~ ~ ONNNX .~13 ~ Ka?rpone ,qnuaa No~ep (9) Knroya (129 zono~KU HoHeo HoHep g) aanucu ~ll~' , qunuHdAa~ ~ HoNep . ~ . ~naarKU ( 7 ) mnoOKU . ~ : ! F: . . . . NoNep ~ � ; ~ , � uunuNdpo � , ' ; ~r ;1'' : ' ~nortC'KI/ �shr ti:' ~ r - Figure 8.15. Recordinq Information on Magnetic Disks of "Counting-Data" Format Key : 1. Index marker 4. Counting 2. Track address 5. Data - 3. Recording Rp-TRACK DESCRIBER 6. Record [Key continued on follow�inq page] 159 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2047/02/09: CIA-RDP82-00850R000504060015-4 rVK VMNN.IAL v.~~ vl~~t [IGey continued from preceding pnqe] : 7. Flaqs 11. Number of recordinq 8. Cylinder number 12. IGey length 9. Head number 13. Leiiqth of data 10. Gyclic verification The beqinninq of all tracks on the w~orkinq surface of the disk is nated with an index marker. The first recordinq on the track is its address. An address oon- sists of seven bytes--one features byte, which indicates the status and use of the track (w~orkinq, defect, main and reserve), four address bytes that indicate the numbers of the cylinder and head and tw~o cyclic verification bytes. The intervals are bytes which are recorded by the control device to delineate the recordinq zones. Intervals Pl and P2 have constant lenqth for each of the types of storaqe devices and the length of interval P3 depends on the length of the previous recordinq. Recording (track describer) with zero ordinal number (Rp) is placed on the track after the track address. The length of this recordinq is equal to 8 bytes and contains informntion of the oontml proqram. The information records of the user are placed on the track after Rp. Ia this case _ the address marker (AM) is initially recorded and then the physical recordinqs of formats--"o~untinq~data" or "countiag-key-data"--are recorded. The countinq zone oontains 11 bytes and includes the followinq fields--features (1 byte), reoording identifier (5 bytes), key lenqth (1 byte), data length (2 bytes) and cyclic verffication (2 bytes). 8.5.4. Maqnetic Dr~.mi External. Storage Devices (NAKB) The magnetic drum is a hollow cylinder, 320-630 mm in diameter and 400-700 ma long, manufactured from dimaqnetic material with a thin layer of ferromagnetic coating up to 40 microns thick. The layer applied to the surface of the drum acts as a magnetic carrier. Magnetic heads, which are arranqed alonq the generatrix of the drum (Figure 8.16), serve to revord informatfon on the drum. The drum is set into rotation by an electric motor at a constant speed. Opon ro- tation of the drum, the magnetic heads create a magnetic field which is used to record infontwtion. The heads record on tracks arranqed around the circumference of the drum. The gap between the heads and the surface of the drum is 20-50 microns. Recording density is equal to 4-8 bits/mn. Movable heads automatically shifted to the necessary track are used in hiqh-capacity drums. The capacity of the storaqe device is determined by the drum dfinensions, number of heads and inforniation recording density. In m~dern computers maqnetic drum stor- age devices have a capacity of 5,000-400,000 bits. The average access time to any section of the drum surface is the time expended on a half revolution of the drum. The magnetic heads are usually arranged in tw~o units. A YeS-5033 (USSR) and YeS- 5035 (Peoples Republic of Bulqaria) type NNID is used in YeS computers. 160 FOR OFF7CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OF'I~7CIAL USE ONLY . . There are eight read/write heads f~r each addressatble track in the YeS-5033 storaqe device. The YeS-5033 has diameter of 450 am~, capacity of 6 l~ytes, 800 tracks on which information is reoorded with clensity of 50 bits/mm, data transmission speed of 1,250 Imytes/s and drun rotation speed of 1,500 rpm. ~ The YeS-5033 storage device cxn be reqarded as a sinql$ cylinder consisting of 100 addressable tracks. There are eiqht write/read heads for each addressable track. The YeS-5033 is used in older models of the Unified Computer S~istem: YeS-1040., YeS-1050 and YeS-1060. ~ ~ . 2 3 Fiqure 8.16. Magnetic Drum: 1--magnetic drum; 2--electric m~tor; 3--magnetic heads The YeS-5035 storaqe device has the follawinq specifications: capacity of 2 NB~ytes, number of tracks per drum of 532, recordinq density of 33.5 bits/n~a, dacta transmission speed 100 Imytes/s and drum rotational speed of 1,500 rpm. Information can be recorded in magnetic drum storaqe devices by three methods; series, parallel and parallel-series. ' with the series method all digits of a number are arranged along a single track of the drums with the parallel method each digit of the number is arranqed on an in- dividual track and with series-parallel the nwnber is divided into groups which are written on the drian sequentially. The most widespread method is the parallel- series. Magnetic drum storage devices, like magnetic disk storage devices, are related to direct access stores. T'he advantages of maqnet~c drum storage devices are large capacity, simple access, capability of multiple readinq and servicing oonvenience. The disadvantaqes are sequential selection and mechanical rotational movement in which the gap betw~een the maqnetic heads and the surface of the dr~ua is retained. The latest m~dels of interchanqeable magnetic disk storage devices, YeS-5061 and Yes-5066, are essentially not inferior in speed to magnetic drum storage devices, but they considerably exceed them in capacity and rapid replacement of the ~ infonaatfon carrier. 161 ~ FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ruK urri~iw~ u~r. u~vLr 8.5.5. Magnetic Card External Storage Devices The Yes-5071 magnetic card storage device is designed to write, store and read large infor~ation files (up to 109 bits). It is used together with the YeS-5551 cvntrol device in YeS-1020, YeS-1030 and YeS-1050 models. The YeS-5071 storage device consists of an intermediate control device and storage device. The intermediate control device provides communiation of the main YeS-5551 control device with the storage device. It is desiqned to c~ntrol information on magnetic cards, information enooding (decoding) and error checking. The storage device is used to store magnetic cards, to move them to th~ read/wri.te heads and to transmit the read information to the control device. It consists of an automatic card file unf t(BAK), automatic card file oontrol (BUAK), card feed unit (BKM) and engineer's console. There is a magazine in the automatic card file unit to store 512 groups of cards consisting of two sections. Up to 32 containers ~ are i.ncluded in the sections. A group of cards is renaved from the selected con- tainer and it is transmitted to the card feed unit by a special hoist. Cards are selected sequentially from a selected group and informati.on is also written and read in the card feed unit. The engineer's oonsole is equipped with manual control devices to check the stor- age device and also for manual replace~nt of card groups in the card file. The YeS-5071 storage device has the following specifications: capacity of 125 Mbytes, re~rding density of 32 bits/mm, read/write speed of 52 I~ytes/s, average information access time of 5,000 ms, capacity of one magnetic card of 2,048 bytes, (16 + 4) tracks on a card, hopper capacity of 128 cards and nnmber of hoppers in storage device of 512. Comparative analysis of the main engineering and functional parameters o.f a magnet- ic tape and direct access storage devices shows that they have approximate].y iden- tical capacity and information exchange speed during writing and reading. The advantage of the direct access storage device is imdoubtedly the short information retrieval time and th~ cost of storing a imit of information on magnetic disks, drums and magnetic cards is ~pproximately an order higher than storing it on maq- . netic tapes. Therefore, these features af storage devices should be taken into account when operating them in computer systems. Chapter 13. User Station System 13.1. General Data User access to computers is gained in TD [~~te data processing] systen~s m~inly through user stations (AP). User stations are peripheral con~uter devices instal- led outside the machine room and connected to communications lines. If a user station contains equipment for manual input of information, it is designated mainly for information gatherinq and if the user station contains a display with keyboard it is used to work in the dialogue m~de. Moreover, a user terminal can - 162 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OFFICIAL USE ONLY - be used autAnomcusly as a da~a preparation device. It may also include teletypes, automatic devices oonnected to machine tools and various installations and so on. But in all cases the user station is the oonnectinq liNc between the oomputer and its usQrs. Information can be prepared and transmitted in a usEr statfom in the ou-line mofle - when data are fed directly fmm devices, keyboards aad d~ocuments to the oommunica- tions channel for transmission to the computer ox in the off-line m4de when data are initially recorded on magnetic tapes, punch cards aad papertapes and then are entered from these carriers. Tn this case the user statioa is equipped with the corresponding input devices. The output data fed from th~ computer to the user texminal can also either directly oontrol specific production processes or be led to intermediate carriers (user stations are equippe8 with punch card and papertape output devices) and for print- out (user stations are equipped with typewri~ers and printers). Al1 the user station equipment is oonnected to the control devic~, which is a com- pulsoxy component of a user station and which performs the following basic functions: establishment and breakinq of ~communications with the MPD [Data transmission multiplexer]; data oonversion from transmission code to the code fo the external devic~e when receiving information from the computert reverse oonversions ~**ien transmittinq information from the user station ta the � computer; calculation and oomparison of check sua~s whe.n transmitting information in blocks; - matchinq input-output and infonaation transmission speeds; provisior~ of autonomous and operatinq modes of f~mctioning of the input-output devices and so on. Besides the control ~vice, the user station also contains terminal devices. Unlike local terminals connected directly to the computer, the terminals installed at user stations are called rempte ~erminals. Alphanumeric tenainals (teletype and typewriter), graph tern~inals tgraph plotters, ELT with graph input) and so on may be used in user stations. Transfer functions are accomplishedby the UPS [Signal shaping device] and modems built into the ~.iser station. _ From the viewpoint of users, u~er statians should meet the following basic requirements : 163 FOR OFF[CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02109: CIA-RDP82-00850R400540060015-4 � rvK urr~~.~ni, u~c ~ivLY the capability of the user w~rkinq in practically any hiqh-level algorithmic language; simplicity of workinq with the system, especially for personnel who are not regular users. The communication lanquage for this interaction should be cloee to natural language and the equipment used should be simple to operat~. accessibility, i.e., the capability of rapid connection of the user to the system at any time; a developed checkinq system that ensures hiqh reliability of data transmissiont ~ short answer waiting t~me= prevention of~unsanctioned access to unauthorized personnel. User stations differ from each other by the types of communications channels, the data transmission modes and methods, the.area of appliation and types of installed te~?inal devices. Thus, for example, user stations may provide information exchange with the oomputer in the dialogue mode when the data are usually entered directly into the computer from a keyboard and in the b~tch transmission nade. In this case the data are first prepared, partially processed and are then transmitted to the computer. Nbst user stations can operate in these twa modes. _ If the user station is not connected permanently to the renate data proc~~sinq system and requires that conne~ction establishment procedures which are pe+r~ormed by automatic output and answering devices, then the capability of oonnection both from the con~uter and fmm the user station is determined by procedures of the - o~mpetition ande~ When several user statfons are connected to the same allocated line, an additional change af control procedure is required to select the type of user station which should transmit or receive data. To do this, the subordina- tion made is used, which is characterized by the determi.ning role of the computer in the communications establishment procedure. Depending on the number of operators working simultaneously, user stations can be individual and group usz type. Group user stations always have a large number af input-output devices and also a storage device consisting of functionally indepen- dent units whose number corresponds to the number of users. The following are distinguished depending on the equipment being used: collective and individual user stations in which part of the work is performed by minicomputers contained in them. These user stations can be used ~n all oper- atinq modes; transmission speed is 9,600 bauds and reliability is 10 . An exam- p~e of this user station may be the AP-50, related to the remote data processing system of the YeS computers; user stations equipped with papertape and punch card input-output devices, = alphanumeric printers and magnetic tape storage devices. The batch processinq 164 FOR (~1FF[C[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED F~R RELEASE: 2007/02/49: CIA-RDP82-00850R040500060015-4 m~de, hiqh speeds (up to 4,800 bauds) and transmission reliability are character- istic of these user stations. The giv~en type of user stations includes, for exam- ple, the AP-4 of the rem~te data prxessing system of the Dnified Computer System= a user station with one or several visual information display devices. Alpha- nwneric display user stations, oonsisting of ELT display devices, a keyboard atic~ printer, are m4st widely used in oomputer networks. Their advantages are high rate of information exchanqe with the computer, capability of editinq data and so on. 'i"hese user stations (these are the AP-61, AP~-62 and so on in the rem~te data processing system of the unified computer system) are used mainly in the dialogue m~de; user stations havinq input-output devices with keyboard. T'his type of user station is rather widespread since it is simple to service and does not place high requirements either on speed or the reliability of transmission. The remote data processing system of the unified oomputer system offers several types of these� user stations: AP-2, AP-3, AP-70 and so on; ~ user stations in which teletypes, characterized by low cost and simple opera- ti~n, are installed. However, the transmission speed in them usually does not exceed 200 bauds; user stations in which there is equipment used to connect abnitoring and mea- suring devices, recording devices, preliminary information gathering and recording devices, bookkeepinq equipment and so o~. 7.'hese user stations (for example, the AP-5 and AP-6 in the rermte data processing system of the tmified computer systean) operate in real time and have rather hiqh speed. Besides the considered user stations, proqram~?able user stations have been devel- oped and used which can partially process information independently of a central co~uter. These user stations contain a processor, internal storaqe and operating system. - User stations are used in the most diverse areas. Thus, for example, user statians that gather data and accomplish real-time control are used extensivelgr in different ASU [Automated control systemJ. Systetns have alr~ady been developed in which data is qathered from sensors or mea.suring devices by means of tenainals. The gathered data are fed ov~er communications lines to networks for subsequent transmission and processing. For real-time control, user stations include either a buffer storaqe or a small computer, sensors, real-time clocks, measuring devices, analog-diqital converters and so on. The aggreqate of these devices is used to observ~e and cor- rect high-speed processes and also to transmit and process the received information. Terminals are used for manufacturing and commercial services mainly in user sta- tions that gather data im m~nufacturinq operations or product output at production enterprises and also for gatherinq data on the exchanqe of goods sold at depart- ment stores and self-service stores. These terminals have built-in data checking or refining devices, automntic devices for readinq encoded marks or taqs on goods or produced products . 165 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00854R004500060015-4 ruK urr~~~w~ u~~, ucvLY User stations operating in the dialogue nade have the highest speeds in rem~te data processing systems. For these purposes the user stations are equipped either with electric typewriters or with alphanumeric or graphic displays. The main disadvan' tage of using these user stations is the absence of oopies of the output results. The composition of a vser station is determined as a function of the wlwnes of in- formation, the type of user problems solved, the infonaation exchange rate and many other factors. 13.2. User Stations of the R~awte Data Processing S~istem of the tAiified Qomputer System User stations of the remote data processing system of the unified computer system ~ contain a wide range of external devices that gather and transmit data, display it on a display screen, enter tasks into the computer, make inquiries and issue lists, nnnitor and manage pmduction and so on. Besides external devices, user stations of the unified computer system are distinquished from each other by the type of communications channels used, the type of connected data transmission multiplexers and somP other characteristics. The basic data on user stations, related to the remnte data processing system of the unified computer system, are presented in Table 13.1. ~ The characteristic feature of the AP-1 user station (YeS-8501) is the presence of papertape and edge-perforated punch card input-output devices, a typewriter, buffer storage with capacity of 160 characters, control unit, telephone apparatus and so on in it. All this equipment ensures semiduplex data exchanqe over conanutated and seqregated telephone and uncommutated telegraph communications~channels at a trans- mission speed up ta 1,200 bauds. when operating in the sutonomous mde, the AP-1 types inforniation on a typewriter with simtulaneous preparation of papertapes and edge-perforated punch cards, prints - out data entered earlier on the papertape and reperforates the tape. The capability of connection to any data transmission multiplexer, the presence of various externa~ devices and the variety of functions performed contribute to the ~ extensive use of the AP-1 in rem~te data processing systems for data 4atherinq, inquiry processing and so on. The AP-2 user station (YeS-8502) also includ~s papertape and edqe-perforated punch card input-output devices, a:Gonsul-260.1 typewriter (YeS-7172), a uadem-200 and so on. The AP-2 has two models differing by exchange algorithms and the type of data transmission multiplexers: one madel is connected to the MPD-3 and the sec- and is connected to the i~D-lA and MPD-2. Both models operate at a speed of 200 bauds. One of the possible configurations of the remote data processing system of the unified computer system using the AP-2 is presented in Figure 13.1. All the enumerated user stations are designed to transmit data in the pack mode. The AP-4 user station (YeS-8504) provides the dialogue mc~de. It can be equipped with papertape and punch card input~-output devices, alphanumeric.printers, mag- netic tape store and a typewriter. A sinqle user statian can contain no more than eight of any eacternal devices ~nstalled at a distance up to 500 meters from 166 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500060015-4 FOR O~iCIAI. U: ' 4 N ~t c( C ^ ~ ~ ~ ~ q ^ ci r c M N M 4MV' ~F ~ ~ ~ ~ C C.' .r. C.' .r. o~D ~ E < ~ g ~ $ ~ ~g`a 'o ~a ` - y` Y~ ~ ~ ~8 8 ~ ~~~g~ _ ~ ~ a~~~^. N aa aaa~ 1p ~ N ~ 7! ~ p'~+ N ~s m'G z x xN ~ae~ ~qNroa,c r~ja'a' ~GM=aTySvl ~ OYV ~C'a~i'SY~W* 1~ v~aY~'~'a{ 11 ~L ! S~ Y S. .+G a~i 4~ LO ~ F-~ ~ lAnwr~~oQ~ a~ ^ ~ oai~ �nuuia*o~t~u N ~ ~ IfWH ~ 'f' ~:~w~meN ~s~~(mxu N + 'F' ~1I11Y ^ ,~pavn~xv o IA O ~ v ~ r1 nid~~o�dau ex ~ ~ ~rCVOYNY OYL~dOAl~i( ~ ~ - . V~ e ~ R1NJIfOmdiU tN ~ 1 ~ T ~ ~ F ~.fYOYRY OYI~tlOtIl~L T ~ U tliMD~CO~d~Y VO~ ~ Wx a N1ClIW~~1i11 a ti ^ ~M Aiaauo~d~u ex W ~ evoona ooaayodia[ j ~ ~ v ad~yo~dau a m ~ ~ ~voa~ waa~odaat b O a ~i ~ aravo~dan ~ y~ + + ' C ~0 v rvoaa oauyodta6 � ~ V y~.~ yaxn~ o dau ~ ~ ~.1 ~ooa~d~ ~ adtyom r. + � ~ rl V ~dau w iM~vo~dau a ~-1 A! eVOw oatayodiaL ~ W a ~ ~~r�uxaoaacx-rr~Aa~ M^~ d' c. ~ 4 a F V tOr1 n U O?0. 0. .~i0. N m -�Z ~a =VaN ~UN~( ~ NZ GC e`o ~1 CC w U Z~ t~~+L� G.. r-1 ~ 'd N lO ~ V~ ~ ~--1 v p~~I CV O C~O v ~ 0~0 ~ C~O N rSa~ LJ,1 {+J [1.1 4tl Lt] ~ f. ~i - ,q i. N l'~! r' ~A t0 ~ ~ ~ �wavo~y - '-C ~ ~ C C C C ~ � 4 4 4 4 4 4 167 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2407/42/09: CIA-RDP82-40850R000500460015-4 rvn vr~�~~.a.+... VJL . ~ . ,t .a ~i 4 4~~'!~ 4 u~ ~ r~ w: ti C ~[ciC ki ~S ti~(CicL . CiGSCtkS ~ C g CC~C C CCCC CCCC E e ~~Q~~~ ~~~~.~.Tn~~~ s ~ N 4~ ..:cj...e+~~N .--~N""N.,- ~^N._:N ~ ~ ~ ' ~~~~s~~~~~ g ~ ~ , ~ ~ , !S - �c ~o - Q `~A. O V O .�r S 8 O C~ ~ Q x op.r~8S ~vopNQg 8 ~o~i.,; r' 8~~8 S U~ ~ONtDN ~~'N"""'N tOCV~^~~ . tDCVtOCV ~ a a ~-a � S x S m~= d= Y S~' I Y S ~s~~ ~n "oa.ssoL GQSQ~ 6~4Y9' ~~V ~~t~. a~=~' /1'Y * ~LY ~YY t(Y X Y OaO~~~~~F ~ ~ m�~~~s:m~~t-o~ aaom~uo~ ~ N N ` , ~ �d:u 'exoiaaco�dau ~ ' IfWH e+~ sMeHmrw r~mSmru N ' `~'f' T , RUT1Y . ~ ~t T Y davn~xtc o ' "E' + p+ + T li10tMO~d~U !N ~ ~ ,'1�~! CPOYR/ Ot1~tlOdt~L � � a � . ~ "4'~ : r wiea~@d~u er k ~ ~ ao : ~voon~ o~uyodu.C � ' ~ ~c.~~~, . . ~ . c) 1NrnedoQdau IM~ - t r. ~ . + ~ . �itdll ~ R1di710~diU ~ ~ N N1N.iL'O~GiU fY , t40~RY OILLJdOdl~~ . ^Y IdQMO~iU ~ � ~ tYO~Y OYl~@OQ1~rC ~ ~ ~ ~ ~ r?~,, j,~ , ~ ~ 1NilfO~i1! ~ ~ ~ �~~'~'~5,~~ ~p 1'. ` ,r~ etro~r oat~yodt~t . . ~ d - .t . ~ tlarD~do~dau ~ ~ a+..i f'i , ~ ~ a0iild7l ~ 1dt710~ ~ i y~ ~ e~l ~ �dau b ira~ro~au a ~�~;f~i O ~ erowo oaa~yod~~t ~ . ~ o a a V ~rue~oao.ier-~esdi7 e+~ V S.~"'~,~ p, ~ U~. V p~p O v U C Z t'" ~ V K7 c~ � p ~ ' ~ ~ c~ ch ~n ~ `,e~ ,n , � YO~ N a~o a~D ~ a~0 a~0 a~D a~G : . a0 +~1 al V V �U,1 W W W W�'~ W W ~ tt] ~7 U t!1 ~ ' ~ tG' tNO~ . . ~ tG E�'d, ~ratcoW " C C C C C C C' C'~. ~ Q 4 4 4 4 a ~G f.~: �4 . 168 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/42/09: CIA-RDP82-00850R000500060015-4 FOR OFFIC[AL USE ONLY t~ ~ - b n ~ I U ~ a~ aa 3 ~ ~ cc a a ^ ~ b _ ~ ^N - m ~ ~i~[ ti~( ~ ~ ' ~f ~p � O~Fi OU g V +~~1 ~ +~~1 ; , oL$ ~ g ^ ~ o ~ ~ ~ ~ uc g ~ ~ ~ ~ ~ $ ~ a � a~ aa ~o+~ ~ : ia~ 'aS.� x o,a,m~xao~s +~~1'~ +dl p C a ~ _ ,~i ~s ^ i~.:=S YIOj,:=~ b ' ~ ,~~~5 ~~~=a ^Y O 0 o ~7ai~3`~o3e~~ ~o V ~i~ V ~ ]~~i~lSOi-Y.L~7?- ~ V ~ QI~ V ( aJou~eu u r ~ l~ 4i ( �~~o~~~~~ - ~ .e c~ a y �d a&aoa~ if W H j r~xriaeM reuutmrn N ~ . . . � . ~O l~ CD G1 O ~SUIIY = ~ YirWNV O ~ Aldl710~I~JY HII ~ I ttOYRY OOl~uO~Il~~ - V < ~C a ~ ~ wauaro�dan n~ ap i F ~ro~rih uat~Uuda~~ ~ u yaenedo~~tao yw - ~ ~ ( -a~da a ntds~dau n ~O ^ evwflwi~ aou~ ta~~ - ~ s�~+ a : i ~ ~ a a ~o~ ~ irara~~aa ~ ~ ~ O ~ epo~a oauyoda~L ~ ~ ~ yownedo~t+u iT O 'Ci ~-i ywarAa ~ 1J~yo� ~ i: ~9 O a .dau M JNiMO~II~Y ~ ~ O ~ ~ ~ ~ L'IIOYY ~w~~n�~,~c _ _ a . ~ o, o 0 O .roiunoao.uw�~ueaa~ M Ja~ ~ a V 1"0 0~0 a~ 'd U VZ~r vTL O O~ m~ ~ v p -cr e+~ 'O ~ pq, a i: ~ i roM N ~ ~ ~ U~ U W'~ C ~ w W w Q � � � � � ~ �-~1 n ~ u$ w~ Gt-pmi+~ au~Fpm ~.-1 N M'~T u1 1raYOw C ~ abaae~"s~00~�y71eq~~+00 p ~ I~ 4 F-c.saat-=x~e[-~c.xop,y~x~e ~ .r 169 FOR OFFIC7AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 ruK utr~~~w~ uaa ucv~Y [lcey cantinued from follaring paqe] : 11. Display 12. Alphaiiumeric printer 13. Zypevriter 14. Mac~etic tape store 15. Pu~ched badqes aad puached nameplates 16. ~ipe of com~aunications chanael 17. Traasmission speed, bauds 18. ~ipe of data transmission multiplexer 19. AP 20. YeS 21. Peoples Republic of Bulgaria 22. Ccmmutated and seqreqated telephone aad uaocmmutated telegra?pli 23. All I~D 24. Version 1, Huaqarian Peoplas Republic 25. Version 2, USSR 26. O~~utated and seqretated telephone 27. USSR 28. Seqreqated telephone 29. German Demxratic Re~ublfc 30. Polish Peaples ltepublic 31. Hunqariaa Peoples ~epublic 32. Seqreqated telephone and teleqraph 33. USSR sad Bulqaria~ Peoples Republic ~ 34. Comautated and seqreqated telephane and segreqated teleqraph 35. Teleqraph appaxatus of five-element co3e 36. CSSR and German Democratic Republic 37. Ca~unutated and segregated teleqraph 38. Teleqraph apparatus of seven-element code ~ An-z . I . ~ ~ ~w ~ ~ I ~ ~ ~ .I ~ rQ~~ ' (9 ~ 1 u�~'~S~'�~ t7) i Mbltiy- eneoaw r M-?AD 9y J 3eM Mnq-? Mon c~~au ~ - = _ , r------ yRC Tr ewdrarN~wd ~r ~NlC 7! 9y ~ I ~r~Pyasx ~ ~ ~4~ ~5j ~ 'o~ ~ ~?sr~wattus~ ~ I ~ ' "qnM~~ ~ra~e~ ~ ~ ~ ~ ~ � Lan_? . Fiqure 13.1. Configuration of Remote Data Prooessinq System of Unified Computer S~stem Using AP-2 (Model 2) [Key on following paqe~ 170 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-04850R000500060015-4 FAR OF't~7CIAL USE ONl.Y [1~Y co~tissued fsom precedin9 Paqel = l. Co�~uter 2. lbdeae-200 3, r_a~utated and seqreqated telephone ooaisunications chaonel 4. Telegraph sigaal coe?vsrsi~ou device 5. Segreqated teleqraph co!~euaications cha:mel 6. Telephoae apparatus 7. Qoatrol device 8. Papertaps aad edqrperforated puach: card input d~eviee 9. Papertape and edqe-perforatad ptmch card output device 10. Zype~rriter the coutr~l u~it (Figure 13.2). The AP-4 also hae a memry ~rith caPacity of 32 lmytes ana ~ operatinq system. it opsrates aaly on co~utated o~a~ications channels through a m4dea~-1200 at a sp~ed of 600, 1,200 and 2,400 bands aad is con- nected to the l~Po-lA, lsPD-2 and M~D-3. - The main desigaation of the PiP-5 user station (YeS-5505) is data qatherinq and also productinq m~nitorinq and oontrol. In this reqard it can be squipped at the desire of users with papertape inputroutput devfoes and v~ to 15 peripheral devfces that oonnect the inquiry read device fs~o~n punched badges and puached namsplates, aa alphanumeric keyboard. ir~quiry dsvioe of ineasured value sensors. Printer and so on. The SP-5 operates over ~c~otnautatad teleqraph oommuc~ications chanasls at a speed of 200, 600 and 1,200 bauds and is oonnected to the 1'!PD-4 , . ~ . ~raenx (6) I . ~ I ~eenn ~ . ~ ~ I 1 ~3~~. ~ Na~uN,rO�v ~ 3err 2 woMan 5 y NMn ~ w I ' 9 ~ I ~ . ~ . ~ ~ny 10 ' Figure 13.2. Oonfiquration of Rem~te Dnta Pmcessing System of Unified O~mputer System Usinq AP-4 Key: l. Ooa~Puter 6. Punch card input-output device 2. Modem-1200 7. Papertape inputroutput device 3. Segreqated telephone comm~mications 8. ~?pewriter channel 9. lrlagnetic tape store 4. Storaqe device 10. Alphauumeric printer ~ 5. Control unft - 171 FOR OFF7CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2047/02/09: CIA-RDP82-00850R000504060015-4 ~'VA V~ /'/~-/AV VJa. V1~L� The charac~eristics of the AP-5 coiacide almost completely with similar parameters of the AP-6 user statiaai (YeS-8506). ~e main differenoes inclu3e broader aomposi- tioa of extert~al devices of the AP-6, qreater amount of oonnected equipmeat and the presenve of a proqraaiminq device havinq up to 16 differeat pr~grams. The AP-11 user station (YeS-8511), whfch pmvides semiduplex c]ata exchanqe m~d~ ov~er seqreqated telephone chanaels at a transmissiou speed of 1,200 a~d 2,400 bauds, includes puach card and papertape input-output devioes, a Konsul-260.1 typeariter, alphanumeric printer and modem-2400. The AP-11 co~na?icates with the oonQutetr throuqh the I~o-lA and MPo-2. The AP-14 user station (YeS-8514), equipped with papertape input-output devices, punch card input device, alphanumeric printer, maqt?etic tape store, display, type- writer and so on, can be used both in dialoque remote data processinq systems artd in batch trarismission systeffi. The ma)ceup of the AP-14 may be variable. Trans- mission is acconQlished over vomm~utated and seqreqated telephone ooamunicatioas channels through a modem-1200 and andenr2400. Z'he AP-31 (YeS-8531) and AP-32 (YeS-8532) user stations vonsist of punch card and papertape devices, magnetic tape store (AP-31) and equipment that uses edge-per- forated cards (AP-32). The ~ransmission speed over ~elephone chaiu?els is 200, 600 and 1,200 bauds. The AP-50 user station (YeS-8550) was developed on the basis of the YeS-1010 mini- con~uter and transmits over conimutated and seqreqated telephone vomaaunications chaanels at a speed of 600, 1,200, 2,40Q, 4,800 aud 9,600 bauds. The AP-50 is connected to the computer throuqh an MPD-lA, 1KP~-2 aad ~iPD-3. The AP-61 (YeS-8561) and AP-62 (YeS-8562): user stations were voustructed on the base of a visual display, by means of which information is exchanged operationally in the dialoque nbde through all types of data transmission multiplexers, except the MPD-3. The AP-61 an~d AP-62 provide semiduplex com~aunications ov~er secTreqated telephone channels at a speed of 1,200 and 2,400 bauds. The AP-62 can also oper- ate over seqregated teleqraph con~~ications cha~els. The main devices of these user stations are: cathode-ray tube (ELT) display measuring 35 cm in the diagonal. The workinq field of the ELT screen perciits display of 960 characters. The methad of print- ing the characters is pointt an alphanumeric keyboard located separately from the display has 28 keys to control the entered information, edit and print it, and also 51 keys of the Rus- sian and Latin alphabets, nwabers and special characters for recording information - in the contzvl device with subsequent display of it on the ELT screenj a typewriter that permits one to type information ooming from the computer or from the display screen. Unlike the AP-61 and AP-62, the AP-62 (YeS-8563) and AP-64 (YeS-8564) user sta- tions include 24 and 16 ELT displays, respectiv~ely, measuring 25 cm in the diaqonal 172 FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02109: CIA-RDP82-00850R400540060015-4 aad vith uorking fi~ld of 240 characters. Ceatralized ca?tsvl of the displays and of their aornespoudinq alphanu~eric ksyboarda is acoosplishedby a qso~ oontrol device havit?q ms~ecry with ca,paaity of 4,096 bytas a:?d located at a distance up to 500 ~estera fran the diaplays. ZM~o typeWrit~ra can also be cannscted to this de- vice. The cozifiqusation of the ismt~ psnoesainq system utilis3nq the l~P-64 is . presented in Fiqure 13.3. i 1 ~ Mo anr�~ N�i ~ r I ~8~i aeH Mna-? ?~ne~r- ~3 n~e- ~ ; I ~ Nan ' . ~ I ~um ~ . u ~ ~ No,u~'~o M ~ ~v ~i ~ ~ I ~4' ~ ~ o- ; 1 , Lsn-as -------------J Ffqure 13.3. donfiquratic~ of Remote Data Processinq Systen of Dnified Computer System Oainq AP-64 1Cey : - 1. Computer 5. Storaqe device 2. Modem-2400 6. Co~?trol unit 3. Seqreqated telephone oomm~nmications 7. ELT display channel 8. Alphanumeric keyboard 4. Typewriter The AP-70 user station (YeS-8570) is desiqned for semiduplex exchaaqe of informa- tion ov~er com~nutated and seqreqated telephone aad uaoomm~utated teleqraph co~umica- tfons channels with a con~puter or other user stations in rem~te pro~essing systems of the ursified oomputer syetem. The transmissiou speed is 100 bauds. The AP-70 is the sfmplext to operate. It fncludes a ltonsul-260.1 typewriter, m~dem-200 or signal~ conversion device. The AP-70 can also be c~nnnected to the aomputer throuqh asy data transmission multiple~oer. The configuration of the data transmission sys- tem usinq the AP-70 is shown in Fiqure 13.4. ' ~ I I PTA ~ ' ~3) j ~6~1 JMM M/!~(�? /~M. ROMM 7f/O~MWY ~~M~ Yy I dl/kAtNNMO 7t/11l~/' NMY KOM00 M~JlI I f5~ ~ l ~ . I . I Lan-m---------~ Figure 13.4. Confiquration of Reaiote Data Transmission System of Unified Qomputer System Using AP-70 [Key on followinq paqe] 173 FOR OFF7CIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000540060015-4 rvn v~ r~~ [Key oontinued from preceding page] 1. Co~uter 4. Teleqraph apparatus 2. M4dem-200 5. Caaitrol tmit 3. c7on~mutated and segreqated telephone 6. ~?pewriter commuaications channel Selection of one or another user station depends on the desiqnation of the rem~te processinq system. Thus, for example, the AP-1, AP-4 anrl AP-11 can be used for data gathering. The AP-61 tA AP-64 user stations, which include displays, are effective in those cases when rapid output of data is required and information must be processed in the computer interaction m~de. The AP-11, AP-14, AP-50 and AP-70 may be used in rem~te data pr~cessinq when the results are fed to the user station. Moreover, the AP-70 can be used in text editinq (for retrieval and output of docu- caents, modification of them and return to the system for subsequent process3ng)� The AP-5 and AP-6 are most effective to monitor and control production. BIBLIOGRAPIiY 1. Materiali X7N s"yezda KPSS" [Proceedinqs of the 25th CPSU Conqress]. 1"bscow. Politizdat, 1976. 2. "Avtomatizirovannyye sistemy upravleniya predpriyatiyami (Met~odY sozdaniya). Spravochnoye posobiye" [Automated Enterprise Manaqement S~istems (Methods of Development), Handbookj, lrbsoow, Fhergiya, 1978. 3. Algoritmy i orqanizatsiya resheniya ekonomicheskikh zadach" [A19orithms and Orqanization of Solvinq Eoonomic Problems], Collection of articles edited by V. M. SaviNcov, No 13, Moscow, Statistika, 1979� 4. Al~yanakh, I. N., "vneshniye zapominayushchiye ustroystva YeS EVM" [F.xternal Storaqe Devices of the Unified Computer System], Moscaw. Sov~etskoYe radio, 1979. 5. Anisimov, B. V. and N. N. Gornets, "Sistemy woda-vYvoda Ts~1M" [Input-Output Systems of Diqital Computers], Nbscow. Mashinostroyeniye, 1977. 6. Burtsev, V. S., "Tendentsii razvitiya vysokoproizv~ditel'nykh sistem i mnogo- protsessornykh vychislitel'nykh kompleksov" (Trends in Development of Highly Productive Systems ar?d Multiprocessor Computer Complexes], Moscow� ITM:i VT AN SSSR, Preprint No ~5, 1977. 7. Grubov, V. I. and V. S. Kirdan, "Spravochnik po EVM i analogovym ustroystvam" (Handbook on Computers and Analog Devices], Kiev, Naukova dumka, 1977. 8. Gural'nik, A. K. and P. 3. Melik-Ogadzhanyan, "Ustroystva pamyati sovremennykh _ i perspektivnykh TsVM (Knnstruktsiya i tekhnoloqiya)" [Storage Devices of Modarn and Future Digital Computers (Design and Technology)],1~bscow� Sovetskoye radio, 1976. 17l~ FOR OFFICIAL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 FOR OF'FICtAL tJSE ONLY 9. Dfvnogorts@V~ G. P, and V. M. Yashin, �SiBtema i appaYaturB obmena informat8- iey v setyakh VTs" [S~?stem and Apparatus of Informatioz~ 8xchanqe ia Networks of C~oaiputer Ceaters Mosco~?, Cvyaz' , 1976. 10. Drozdov, Ye. A., V. A. Ramarnitskiy and A. P. Pyatibratov, "MaoqoproqrammaYyo tsifsovyye vychislitel'nyye mashiny" [Multiproqram Digital Camputers], I~bscow, Voyenaaye izdatel'stao, 1974. 11. Drozd~ov, Ye. A., V. A. 1Camarnitsaciy and A. P. Pyatibratov, "Elekts+onnyye vychislitel'nyye mashiny Yedinoy sistemy" [Computers of the tlnified Ccmputer S~rstem] . 1~bsca?, l~lashinostroyeniye, 1976. 12. Davis, D. ar?d D. Barber, "Seti svyazi dlya vychislitel'nykh ~ashin" [Conmuai- cations Networks for Coaiputers~ , Irbscow, Mir, 1976. 13. "Yedinaya sistema EVM" [The Lfiified Computer S~rstem], edited by A. M. Larion- ov, Moscvw, Statistika, 1974. 14. Zhuravlev, Yu. P. and Yu. S. ~kin'shin, "Sistemnoye proyektirovaniye pamyati TsVM" [S~?stems Desiqn of Digital Ooaguter Memories), Mosoow, Sovetskoye radio, 1976. 15. Kagan, B. M. and M. M. Kaaevskiy, "Tsifrovyye vych3slitel'nyye mashiny i sistemy" [Diqital Computers and Systems], Mcscaw, Enerqiya, 1979. 16. Kartsev, M. A., "Arkhitektura tsifrovykh vychislitel'nykh mashin" [donfiqura- tion of Digital Computexsl, Nbscow, Nauka, 1978. 17. "Kataloq. Yedinaya sistema elektronnykh vychislitel'nykh mashin" [Cataloq. The Unified Computer S~stem], Vol 1, lYbscow, Stroyizdat, 1979. 18. Karolev, L. N., "Struktura EVM i ikh matematicheskoye obespecheniye" [The Structure of CoaQuters and Their Softwarel, P'sosc~w, Nauka, 1978. 19. Mayorov, S. A. and G. I. Novikov, "Struktura elektronnykh vychislitel'nykh mashin" [The Structure of Computers], Leninqrad, Mashinostroyeniye, 1979. 20. Martin, J., "Seti svyazi i EVM" [Do~ua3-cations Networks and Computers), Parts - 1 and 2, Mcsoow, SWysz', 1975. 21. Martin, J., "Sistemnyy analiz peredachi dannykh" [Systems Analysis of Data TransmissionJ, Parts 1 and 2, Moscow, Mir, 1975. 22. "Mul'tiprotse~somyye sistemy i parallel'nyye vychisleniya" [Multiprocessox Systems and Parallel Calculations], edited by F. G. Ansdow, translated from Enqlish, Moscaw, Mir, 1976. 23. "Osnovy postroyeniya bol'shikh informatsionno-vychislitel'nykh setey" [F?snda- mentals of Designing Larqe Information-Computer Netw~orks], edited by D. G. Zhimerin and V. I. Maksimenko, Moscow, Statistika, 1976. 175 FOR OFFICUIL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4 APPROVED FOR RELEASE: 2007/02109: CIA-RDP82-00850R400540060015-4 l V~\ Vl ~'~~.lA~r VJ~.. V3\L+� 24. Pozin, I. L. and V. K. Shcherbo, "Teleobrabotka dannYkh v avtomatizi-r�vannykh ~ sistemakh" [Remote Data Processing in Automated Systems], l~loscow, Statistika, 1978. 25. "Protsessor EVM YeS-1020" [The Processor of the YeS-1020 Computerl, edited by A. M. Larionov, Moscow, Statistika, 1975. � 26. Pyatibratov, A. P., "Vychislitel'nyye sistemy s distantsionnym dostupom" [Computer Systems With Remote AccessJ, Moscaw, F.herqiya, 1979. 27. Samofalov, K. G. and G. M. Lutskiy, "Struktury i organizatsii funktsioniro- vaniya EvM i sistem" [Structures and Functional Organization of Computers and Systems], Kiev, V~?sshaya shkola, 1978. � 28. "Seti EVM" [Computer Networks~, edited by Academician V. M. Glushkov, Mascow, Svyaz', 1977. 2y. Sist~na dokumentatsii Yec~inoy sistemy EVM" [Documentation S~?stem of Unified Computer system~, edited by A. M. Larionov, Moscow, Sta~istika, 1975. 30. "Sistema matematicheskogo obespecheniya YeS EVM" [Software System of Unified Con~uter System~, Moscow, Statistika, 1974. 31. "Sistemy peredachi dannykh i seti EVM IData Transmission S~?stea~s and Canputer Netwvrks], edited by P. Green and R. Lackey, Moscow, Mir, 1974. = 32. "Report of Central Statistical Adminis~ration, USSR, 'On the Results of F'ul- filling the State Plan for Economia ar.d Social Development of the USSR in 1978 PRAVDA, 2U January 1979. 33. "Spravochnik razrabotchika ASU" [Handboak for the Developer of Automatic Control Systems], Moscow, Ekonomika, 1978. 34. Tanenbaum, E., "Mnogourovnevaya orqanizatsiya EVM" [Multilevel Organization of Computers], Moscow, Mir, 1979. 35. Flores, A., "Orqanizatsiya vychislitel'nykh a?ashin" [Organization of Compu- tersl, translated from English, Moscow, Mir, 1972. 36. Flores, A., "Vneshniye ustroystva EVM" [External Devices of Computers~, translated from Enqlish, Moscow, Mir, 1977. 37. Shelikhov, A. A., "Yedinaya sistem~ EVM" [The Unified Canputer System], Mos' cow, Energiya, 19%5. 38. "Elektronnaya vychislitel'naya ma~shina YeS-1020" [~e YeS-1020 Computer], edited by A. M. Larionov, Moscow, Statistika, 1975. 39. "Elektronnaya vychislitel'naya mashina YeS-1030" [The YeS-1030 Computer), edited by A. M. Larionov, Moscow, Statistika, 1977. 40. "Elektronnaya vychislitel'naya mashina YeS-1050" [The YeS-1050 Cbmputer~, - edited by A. M. Larionov, Moscow, Statis+tika, 1976. COPYRIGHT: Izd~~el'stvo "Statistika", 1�8~ 6521 ~D CSO: 8144/0540-A 176 FOR OFFIC[AL USE ONLY APPROVED FOR RELEASE: 2007/02/09: CIA-RDP82-00850R000500060015-4