YW
LMrectofite of iBtrUlsffKt
Soviet Bloc Computers: Direct Descendants of Western Technology
A Rraeardi Psper
This paper was prepared
jf Office of Scientific andCommeou and on cries are wekomebedrjected
.OSWR.O.
solicit bloc computers: direct descendants of western technology
Sonet Bloc will continue to rely heavily on Western computers
o guide (heir Indigenous computer developmeiii efforts
10through acquisition of Western hardware and icchnol-
ogy, for the lack of indigenous production capabilities.'
The Soviet Union and Eastern Europe lag the West in all aspects of genera] purpose digital computer technology, ranging from at leaat five yean for nuciocornputers (also known as personal computers or PCs) to more thanears io high-performance magnetic disk peripherals. If production volume and reliability were taken into account, the lag would be even greater. The progress that has been achieved b, in largeetail of acquiring Western technology, ranging from micrordectronk and printed circuit board prodoction technology to use of Western computers forg
The Soviet lag in computer technology and production isleeting military and civilian computern militarythe Soviet* are avoiding complex multimissioei systems that require high-performance computers and disk drives. The impact is most critical for those applications with intensive computtliooal requirements, such aa ballistic missile defense, and those with large data bases, such as command and control. Without significant advances in computer technology. Ihe performance and reliability of future Soviet weapon systems could be jeopardized. For example. Moscow's ability to develop or acquire advanced computer technology willirect impact oa iu ability to deploy improved ant ibal lis tic missile (ABM) radars for reentry vehidcand multiple target handling. Soviet managers and military leaders have become increasingly conversant with tee role that computer-aided design (CAD) can have on improving the avanabffity. reliability.performance of weapon systems while reducing tbe human and material resourceoviet capability to design and produce composite materials for the next generation of airborne and space vehicles depend*on modem computers in the laboratory and on the plant floor.
In civilian developments, plans for Industrial mcdetwatlon will beby the lack or up-to-date computer systems. Lack of engineering workstations and high-speed computers will continue to stifle theand creativity of Soviet scientists and engineers. Engineeringare needed to implement computer-aidedas
9
CAD, computer-aidcd manufacturingnd computer-aided (citingare key dements in moclernlzalioo of tbe Sovietand alsoviation industry
Implementing increased CAD usage will require not only eogiriecring workstations butomputer literate work force that is able to takeof ibem. Without large numbers of computers for education, the Soviets cannot implement their plans for extensive training program. to prepare engineers, designers, and tech nicians working in CAD. CAM, and computer-aided engirtceringt wdl as cr mpateriied management information systems. These programs will require significant development efforts and substantial improvements in the quantity and quality of production of Soviet computer equipment
Soviet Bloc PCs are produced in small numbers, are low-performance systems, and are generally designed for special purposes rather than for general office or laboratory use. Even (be limited numbers of PCs produced in the Soviet Bloc are heavily dependent on Western-manufactured Components. Soviet plans to produce and makeillion general purpose PCs by the end0 are. in our judgment, unrealistic.
The Soviets and East Europeans arc six toears behind the Westit minicomputer technology andoears behind the Westbit superminicomputers In the important category of large mainframe computers, the Soviets are having great difficulty evenmodels equivalent to those available in the West eight toears ago, and this gap Is growing. We estimate that Soviel development ofsdentific computers (oommonly grouped as supercomputers) Lags the West by more thanears and will probably remainoean behind through theecent organizational changes anddesigned lo improve Indigenous computing capabilities have had little effect to date, and we project tbe gap will grow through
In our view, the Soviet bureaucracy will find it difficult, despite Mikhail Gorbachev's modernization program, to take the necessary steps to correct many of ihc computer industry'srecognised problems. If the Soviets obtain turnkey production facilities or detailed production know-how from
thethey have done in themay be able to narrow, at leastpecific technology gap. The proliferation ofcomputers in the West will increase Soviet prospects for acquisition, potentially allowing them to decrease their lag with the West However, even with turnkey facilities and increased acquisition of Western hardware, the best the Soviets can hope for over the nextears is to slow the ra'-at which the Western lead on computers Is growing
Contents
Slate of Computet HardwmTedntofcig
Systems
Kroeocnpu lers
arid Surierminicom outers
Computers
SdcntifK. Computers
DuV Drives
Disks
Tape Drives
the West for Advanced Hardware
Bloc Needs
Technology
Circuits
Printed Circuit Boards
Manufacturing Technology
Computer Servicing
Drive Production Technology
estetn Sources of Supply
Approaches to Acquisitions
for Civilian and Military Programs
of Computer Parameters
USSR-.SeJectrd Microcomputers
CTwhosloiBkia: Selected Mtctwnvputers
East Germany: Selected Microcomputers
Hungary: Selected Microcomputers
Poland: Selected Mrcrocomputers
Bulgaria: Selected Microcompeters
Romania: Selected Microcomputers
8.
Europe: Minicomputers and Superminicomputers
Europe: Ryad Mainframe Computers
Highspeed Scientific Computers
Europe: Amy Processors
Europe: Rigid-Media Magnetic Dak Drives
Europe: Floppy Disk Drives
Europe: Magnetic Disk Packs
Europe: Reel-lc-Red Magnetic Tape Drive*
Europe: Cassette Tape Drives
Discussion ol Slate of Soviet Computing
Disk Drive Technology at Sura Zagora
soviel bloc computers: direct descendants of western technology'
The USSR has made aubstautial progress derin< thein computer toclinology and production icchniquca, largely because of Moscow's heavy ieli-attce oa ibc West- Despite tbe USSR's large-scale acquUitions, Its comparative progress hai been over-whelmed by rapid advances made ia tbe West and in Japan.m ma hits the Soviets* italui rela-tive to Ibe West in several important ircas oftechnology. These estimates are based on the difference between the Initial delivery dates of fuee-tiona my eejuiraleot US and Soviet civilian computer prod licit."
There arc many reasons why the Soviet! trail the West (incltiding Japan) lo computer technology;
The centrally phoned Soviet economy does not permit adequate flexibility to respond to design or manufacturing changes frequently encountered In computer productico. This often resultshen, age of critical components, especially for new products.
I he. inordinary compaat ion of infortr.ai.vi in (heon (ethnologies withmilitaryeatricii the Dow of information This results in much depbcatlon of work becauseack of knowledge about other activities,
The Soviets are preoccupied with meetingquotas, frequently at the eipense of component and system quality control.
Thereack of adequate Incentives for Soviel managers to take (he risks associated with innova-troni or new technology.
1 There is poor coordination between design institutes aad prod action facilities. This problem has resulted in products that have to be re-desijtned toactory's production capabilities.
The Soviets las in computer-aided design andmaoufaeturinx (CAD/CAM)tacauteate development start and. Ironically, because of the lag io computer
mi ill.
> Soviet officials are concernedomputer couldowerful tool for aotirt-roluliceiary activity androlifcraijoc of coenputera might reduce the tight control of information la (he USSR. These coneons have delayed aod restricted access lo sad firsthand knowledge about computers and their application!.
There are provincial disputes within aod between ministerial and Institutional organizations.
Poor cut teener support iadudingfeedback, poor installation support, and delayedresult In reducedand prodtictiviiy for computers In use
Similar reasons account for the Soviet lag intechnology, automation, instrumentslion, and testoem Lags in comrooerit* aod basic electronic tools that are essential for modernalso contributeo the Soviet lag ia computer technology
Although (he Soviets have hid sufficient numbers of computers for high-priority, low-volume military aod oiviliao pro/ecu, the rctnatiung user community.Soviet Bloc allies, has experienced shortages and delays Inesired computer systems. The ihorUie ofarnation equipment hat seriously hindered the modem it* lion of Ihe Soviett rial base and. thus, the growth of lis economy.
Iliitcetcally, there hasendency In the USSR to avoid the complex mutlimiuacei militaryfor which computers are an essential
are frequently preferred io Ihe Wesi. The generally conservative Soviet weapons design philosophy has probably not taxed Soviet computer capabilities in the past. Also, Ibe large physical sire of Soviet computers has probably discouraged their incorporation into weapons systems (see figure 2)
Lags in high-speed scientific computers and in high-performance magnetic disk technology haveSoviei computer system performance for applications requiring high maul-output data rates, such as large real-lime command, control, andsystems. These problems have beenacknowledged by the Soviets in their writings and
in diicussions with Western colleagues. Indeed, early computer inadequacies ledoviei decision in tbeo copy US computer technology
Tbe Slate of Computer Hardware Technology
Major Systems
MicroeompuItrt. Microcomputers available in the USSR arc of low performance; produced in small numbers; and generally designed for special purposes, such as machine-tool control, process control, and
RouHdtabU Discussion of the State of Seeiet Con am Int
an extraordinary rcaindlable diicussionin an7 Soviet open-press article, high-level Soviet rdhclalt gave irveral glaringaf proolrms with Soviet computer production, capabilities, and usage. Participants Included Ye. Vrlikhov. xrice preside* of the USSR Academy of Sciences; I. Bukreyev. lint deputy chairman af the State Committee am Computer Technology andUs: G. Ryabov. chief designer of supercomputers: B. Yermoiaytv. deputy thief designer of the ES {Edtnaya Syitema or Unified System! seriesof mainframe computers: and V. Kurochkln. first deputy minister af the radio Industry
The roundtobie discussed the urgent need to Increase ihe mean time to failure (hi TIF, one measureomputer's reliability/ af Soviet computeri.assured hit colleagues that3 the6 and1 mainframe computers produced by his Industry would have iheir iXfTTF increasedours las compared with typical hfTTFsfor Westernours) However, even these minimal gains comeigh cost to therevious Increase In the UTTF of Ihe1 from ISOourswo-year effort, SSO squat meters of additional manufacturingdditional pieces of production and test equipmentostilliondditional workers,ore production time. Kurochkm emphasised ihe resistance encountered to these types ofthe factory ntonoger's viewpoint they ore completely nonproductive expenditures. Into ike kith cost af these measures, they leadecrease in production volume
The participants agreed that the computationalof Soviet computers art limited by Soviet component industries, primarily semiconductors and printed circuitChich substantially lag
'ii ike Puailnn worn1 far tnirt. Is often use.
Wesiern Induitries. This leg Is largely blamed on Soviet failure io make the capital Investment when It was needed. Veltkhov staled thai the output of the Soviet electronic nuuhlne-buildtng Industry will have to Increase fivefold5 fust to maintain the gap with the West. Soviet Minister V. Kolesnlkov pointed out that the practiceesigner simplyestern computer, circuit for circuit. and expecting Mintlektrtrnoprom to develop the circuits Is no longer feasible. The panelists agreed that developmeni of the next generation of Soviet computers will require doge cooperation between the various Industries andthat take Into account the lack of Soviet In large-scale Iniegrailon (LSI) and very-large-scale Integration (VLSI) ctrctitU.
The final topic of discussion was the effectiveness of computer usage In the Soviet Union. Quoted stallsllcs showed lhat. on average, large malnfrcmes that are expected to be useday were only being usedours and,orst case, only about seven hours In some ministries. According toV. Bezrukov, director of the USSR Cosplan computerariety of reasons contribute to this uruleniti'liatlon. Including power fluctuations that cause computer malfunctions, computers that are not Installedcf lack of false flooring and other Installation material, shortages of disk drives (onlyegabyte drives are allotted per computer) thai cause computers to stand Idle part of each day, and ihe lack of basic supplies, such as printer ribbons. S. Bushev. director af the USSR Central Statisticaladded that, although all regionalhave computers capable of communicating with each other, lack iff hardware to connect theto ihe communications lines Is forcing data transfer via telegraph, with subsequent manual data entry. The point was made that If these problems occur In the largest, best equipped computer centers, the ordinary user must suffer even worse
measurement/test control. Thef which perhaps several thousand are(he onlybit system thai is in wide use ia the USSR. Upgraded versions of ibeesignatedX. arc now being delivered AD of these, however, are designed for industrial-typelions. Wc do not believe tbey arc designed for desktop or "PC-like" applicaiions thai could satisfy the Soviet requlrementow-oost.general purpose, microcomputer system.
Beginning in tbe, the USSR embarked on several programs lo mass-produce PCs to satisfymu ml;
The Soviet AOAT (seeodeled ifier ihe Apple II, was to be ihe primary machine used in ihe computer literacy program and was widelyto primary and secondary schools. According to many Western experts, the AGATotableota! ofnits were producedhiee- to four-year period. The Soviets eaperi-enccd severe problems in hardware reliability and produtibility and in software availability.
> The Bleklromkaas introduced aftei the AGAT ami was, we believe, somewhat moretotal productionate Is estimated at
fewer0 unns. One significant design feature was lis partial compatibility with US Digital Equipment Corporation's (DEC) machines.echnical evaluation of tbe
' suggested that,the computer ii being sold publicly, il is little moteoy, useful only for games lad simple instructional applications.
The first Soviet copy of tbe IBM PC, the. was to be the principal tool for scientists and engineers and was lo be widely distributed at the tniversiiy level Production of Ibes limnedewearroduction of newer copses ofsee fcgurehe ES-IS4I. Ibend others. These machine* representprogressive effort by ibe Soviets lo standard! the industry on the IBM PC, which will enable tbe Soviets to take advantage of the wide variety of compatible hardware and toft-ware sold on the world market.
Successful miss production of PCs Is highlyon the availability of supporting technologies.
notably mlctaoiocetso** 'bins
Jthereeneral shortage of Integrated
circuit*nd large percentiles ol* thosedefective.} tbe Soviets cUimcdof theiaoprocessor, which6 sited ta an earl* model of the IBMproduction
volume s> limited too mat the reliability of theooe. Ia addition, adequate supplies of rugh-capscr-ty memory devices (for eiynamicaccessare limited, with production only ia the hundreds of thousands per year, as compared with hundreds of million* per year in the West and Japan
avc claimed that tbe USSR plans to rely on liatlem Europe lo augment Soviet production, but ibese countries cannot lalisfy their own internal demands. Despite Moscow's attempts to direct and coordinate microcomputer developments among Eastern European countries,haotic At though eachnique ind our technical information it largely limited to claims made in Soviet Bloc open literature and brochures. East Bloc microcorriputer production can begCLeraUy by:
A wide variety of different models being producedarge number of producers using handThis results in rstretnety limited qualities,ew hundred or less.
A heavy reliance on critical components (formicroprocessors and memory devices) from the West. This means many "Soviei Bloc PCs" are being assembled from imported Western kits.
A heavy reliance on limited and uncertain supplies of Soviet electronic component! of questionable reliability.
An almost total icltsncc on Westernperipherals (such as hard disk drives]
ttempfiag lo concentrate prodaciionelatively small number of models
total production to date has been lessnits. Until recently the Bulgarians had not used IBM-compatibleand,esult, their products found little acceptance outside of Bulgaria like other Com muni si countries, they now produce mainly IBM PC-cornpalibie mod-eb, using many critical parts illegally imported from the West (see figure S)
Eastbe most country, ai leastechnical point of view.oncentratedmall number of models, the industry is concentrated atsuccessful producer of largesystems are ollercd in modern configurations. PCs Ibat are available reportedly are highly regarded and arc reliable. There b,roblem of availability, because tbe products are based on East German or imported Soviet microprocessors ibat are In short supply. Total production to date, at in the case of Bulgaria, is on,ihe ordernits C
. (see figure
6|
According to open literature, Czechoslovakia isserious parts supply tod production problems and produces only minimally acceptable microcomputers in verynits
this< bio )
. Im Gurnet Pebolron EC IIS' PC
and Pc4and have made progress InIBM rf* tjtli machines, using criticalmainly from the Fir East (sec AgereHungary and
Poland have been quite successful in acquiring-and unembargoed complete Western PCs,IftMs, despite Coordinating Committee on Export Control (COCOM) regulations. At the otber extreme, Romania seems to be insignificant in this technology and has no plans lo expand its limited capability. All Bloc offerings suffer fromlack of availability of either domestically produced orhard disk drives.
feri amd Stp'tmiHitomtuitrt. la our
judgment, ibe Soviets and East Europeans arc six toears behind the West inbit mint-computer technology andoears behindbit superminicomputer technology, Tbe Soviets are now producing minicomputersire andthat were available sn tbe West nearlyears ago. Moreover, they did not deliver their fustuntilears after the first US model.hows the lag between initial Western
productionomputer and Soviet Bloc productionunctionalhis Soviet technology lag is translating into an even greater applications lag. We estimate that the USSR is at leastears behind the West in numerical control and flexible automation systems in manufacturing, which depend heavily on embedded minicomputers.
In the, Soviet computer-related literature and conferences began to reflect increased activity in minicomputer technology and applications. There hasteadily increasiBg number of references ia Soviet liieralurc to minicomputer applications for CAD and CAM, including process control.esultuccessful Council for Mutual Economic(CEM A) cooperative program in development of mainframe computers, the Council of Principalof Minicomputer Systems was4 to coordinate minicomputer development within CP.MA. The Soviet Union assumed the major role and developed four new minicomputers: ihehe
-ivee-ci
Figure I
Western lxnd Over Soviet Blx In Mlnl/Supffmlnlcaraputcri
ofproduction
It re olt-impmtk
* Ma*
i HI
asssaa
VAX ll'TK
w U
ISWJ
I'll ol Wnum iiiiiiii FiDdueiienonMutllnoIlk* dwi
ihend theToilnd proosMy the SM-S. ctmutcta the tint generation of SM rmnKueiiputers. Ibeamily.ote lypical configurations of Soviet minicompgltrs )
Thendre modeled after tbe US Hewlett-Packardrchi lecture and prima nly uted forontrol. Thend there modeled alter the low cad of tbe DECinicomputer line and are interfiled primarily for iclentific and cngincerino conlrol
itheSM.iandpparrnlly can execute moat DEC aoftwaic -ithoul modification
odeli arc betes produced in tberana. Caechealewa'ca. and
3 that other computer models from Eail Gertaaay and Romania an la theiass or are being nwdeled arte* theeries.
In the, according to openA countries began coordinating production of the second generation of SM small computers that Includes both microcomputers and minicomputers, the SM-II family. CEMA bas adopted new nomenclature for thtt next generation:
SM-JO; General puipose microcomputer* forsuch as control and communication a.
t-iU<f4tiir-t I
omputers maintaining software compatibility with ibeamily.
igh-performance, real-time small comput-ers; Ibis class wDI eventuallyii superminicomputers.
: Special processors for character recogni-lion, fasttransforms, and other srxciatticd applications, .
Nearlyystems using this nomenclature have been identified In Soviet open literature. Many of the systems are still under development; other systems seem to haveew name for an old system. We believe the lack of successful production of most of these models indicates that iV Soviets arc falling further behind Ihe West
Sowci BJoc countries are developingbit supeminicompuir.il.Hungary, and (lie USSR arcio copy ibe VAX II/7S0 TheHungary, and East Germany0 copy (ice futuresOn the, basis of our limited technical datamachines, wc believe tbeif performance islower than thai of the DEC originals.German0 copy,have ocbe memoryVuhaj point
over, tbe Soviet copy of the VAX II/7J0 is much ltf|er ibannd requiresimes the ckcirical power.
the majority of
nes" arc, al best, only ia ihe prototype or very limited production stage and thai it will be stveril yean before there it substantial production of aay ofopies East Germany seems io be the mosi successful so far inax copy
7
fait, a lifted visitor was able lo view ihtthe computer and observed eiCse ofelectronics in eiact cbip-fot chip and board*copy of the DEC original Some of thewere copied from those of Westernmanufactuicrs rather (ban directlyIn addition. Western dish drives arewith an the machines
llie Sonet Iliac is In ihe process of developing copies of (he newerbit superminicomputers. The So-ietn Germs cy. and Bulgaria arei rig io copy the DEC HicroVAX II. These copies are scheduled for availability1 (
. jTlie Research Center of the Microelectronic Combine (KME) in Erfud, East Germany, started lo reverse-engineer theCsompletely new semiconductor production line was being developed by Zeissccommodate Ibis effort.
iheie VAX copies probably male eUensive uie of imported Western parti.mple. when the East Germans displayedprint trade
,Strrei_
ixx
East Germany alio hai copies of ihc DEC VAX 8XXX series under development al Robotron. The work is reportedly at the design stage add will require the production of several specialiiedort on developing0 aod the0 is progressing ia parallel We believe itbe at leastears before these machines reach serial production in the Soviet Bloc. Duplicating these computers will require East German advances in printed circuit boards,mcmof ics, ipplication specincnd overall manufacturing capabilities.
Computm. We estimate that Ihe Soviets
arc eight loears behind (be West in development of general purr*kc mainframes (secn the West has been so substantial thst the Soviets arc having technical difneuttict even holding ground in this area. Despite pcoeoonccnacau to the contrary. Soviet developments represent only margie-al cvolutio lary movement. Soviet efforts to introduce Ryad III, their newest family of computers, have been delayed
ill representarginal improvement over Ryad II.believe the Soviets will continue to fall even further behind the West over the nest five toears oceanic of conttnuinf problems with bringing new machine* into production, unreliability of machines ihey do prod ace. and lack of peripheral support. (Figure IS comparer the operating ipeed, in millions of instructer second IMIPSL of lop-of-the-line US and Soviet mainframe computers0,
Soviet Ryad mainframe computers are modeled after IBM systems- (hen Ibe; the Ryad II on tbeexcept the largest model, which may he based on tbe; and the Ryad III ostensibly on theX and IbcX models. Each of these Ryad series consist of several models that vary in their computing capabilities and arc intended to aaiiifyspecific dasa of user. These mainframes ne widely used in military and civilian application!
odels ate no longer produced, although most models are still etterurvdy tsscd throughout the Soviet Bloc Although modeled after (heeries, we believe the Srmctseries of evolutionary upgrades beforechieved significantwith the IBM. The high-end machines in (he aeries, partlculitly (heuffered serious delayi in delivciy and went into production only after introduction of the Ryad II. Peripheral development, ia particular duk drives, substantially laggeddevclcproent, further diminishing the utility of already marginal computers. The greatest problem with theot its lack of reliability. An article in Pravda described inadequatdy trained servicelack of spate parts, lack of automated test equipment, and shoddy service as common problems.
il
Figure 14
Western Old Over Bloc In Mainframe Cat
iv
or Bloc
pindimitoni" ii v-
m
p'oduedan
meat significant icrntnplUhcocai of Ukeries dwloprnent was the aubJiihnicotom-patiblc Udc of computer system* foe Soviet Bloc production
upgrades before aclsseving technical comps ability -iih tbecries. Fiftueso ISariety of Ryadodcli in typieal configurations.
amily is In production and isufficient numbers to satisfy s( least high-priority users Although the Ryadodels offerover the Ryadas virtual mernory. larger disk stores, semiconductor memory, and aa expanded Instructionsuffer many of the same reliability problems as the Ryad I. partkularly in Ihe larger machines. In addition, the lack of high-capacity. hlgh-reliabUity disk drives continue tocomputational capabilities Al with tbeodels, the Ryadodelseries of
The Soviets ore hiving difficulties with Ryad III development. Tbe Ryad III program, announceds based on faster, more densely packed logic and rjsemory circuits, larger and more advancedstores, improved telecommunicationsand specialized processors. However, we believe delays and problems with this program may be(he Soviets to classify upgraded Ryad III as Ryad
Ills. Die lew models thai have been describedliterature or viewed by Western observersseem lo meet the outlined teals ef the6
(seedisplayed7 trade fair asa Ryad III model did not seem toubstantial upgradeyad II. Although the Soviets claimed it operatedpeed ofIPS.C valuated its operation al closerIP (approximately com-ratable. The Hast Germans(heir Ryad III7 alB Plodivfair in Bulgaria. Accordingress article, it has an operating speedIP
J
if ihe Soviets continues nt't evolutionary path, we would expect Ryad Ills to be based on the IBM <JXX serieseries of small mi inflames first produced by IBMl the low end ind on the IBM JOJX at (be high cod.
J3-lhe East German Ryad model EST Oil is supposed to be based on tbeX
According to public announcements by tbeRysd developments will include upgradesRyad III and developmentyad IVthe basis of past Soviei performance,III models will not be in servicearliest. The limited information we haveeries suggests that it is under(hai Ihe Soviets cipect lo place ii in service.
Germany's model of ihe Ryad IV would copy IBM's "lalestse complementary metal oxide semiconductor (CMOS) technology, and haveinIPSrototype is scheduled for1
If ihe Soviets and (heir allies continue to follow IBM's lead, they will need Io do extensive reworking of tbe IBM design to adapt it to available Soviet technology. Complex mullilaycicd boards, dense packaging of components, and sophisticated cooling technology thai IBM began using0 forX models are beyond Soviei capabilities and are likely lo remain so through tbe. In addition, ihese cornputer performance improvements will require Ihe Soviets lo make parallel advances in the development ofand itoragc technology
Iligh-Sfittfi SclfAiifie Comfuitn. Sovieiof general purpose, high-speed scientific com-ikjteit (commonly grouped togetherequired for large-scale scientificthe West byo IS years and will probably icmalno IS years behind through the
i
M
Hl*rt ItKyal II iyafm
Sarin ESlfHa mtlrfrtmn. PyaJ In
owever, Soviet research on computerthat use large-scale -parallelism" could aUow the Soviets to significantly uKrcase iheir eapa-biliiies for certain applications. We believe that, ai present, the Soviets have no machines in the true "supercomputer class. The best Soviet scienlific computers have only one-sixth the speed of thdr Western counterparts (sen, and Soviet claimed computer capabilities are probably greatly exaggerated
Strict export controls on supercomputerprecluded the Soviets from developing abasedestern system. Left todevices, the Soviets have not acMcvedcapabilities using indigenous
receoily
called Soviei efforts lo develop supercomputersfailure and claimed that another five orwill be required before indigenousto the ilt be"he tackupercomputer
has already curtailed several projects ai ihe Institute of Space Research
Theirst introduceds stillused by the Soviet sdenliSc computing esiab-lishmenis. despite its limited capabilities ll was prigj-nally designed withilobytes (KB) of memory and used magnetic drums for mass storage. It was updated in ihe, its memory wns extended to
B. and it used disk storage. Calculations of ju speed put ii ioillion floating point operations per second (MFLOPS) range. This performance level, which decreases to5 MFLOPS when input and output of data to secondary memory is required, is comparable to many advaoccd Western PCs.of theoded In the
8 ihe developers of thennouncedmachine, the El'brus. The Ei'brus-Idesigned to achieve high speed by using one.orrtang stfrrulianeously onproblem. Hy ^
perhapsingle processor models had been delivered to scientific institutes and lhat the Soviets planned lo produce them at the rate of roughlyer year. However, many of theomputers were not working reliably and. In someere not even functional because of missing subsystems
Theollow-on appears to be idenljcal in architecture to ihe Horus-l. with the primary differ, ence being ihe level of technology used. Along with an overall upgrade in ihe sophistication of senxlcooduc-tors used, theas additional vector rnocess-ing components. Thelso has from One torocessors and, by Soviet claims, is intendedIPS
Fltrut. 20
Corn part ion af US and Sorterpalm
h
ll'brvil
cmv-i
iirfrfoonuke claim otunlom tpce* incmi.
in ill" liltivi'l
MtrrSuk. President of the Soviet Academy of ScKncei iiaied
hat theould be available for delivery lo Soviet industry and researchbyvidence thatbrine delivered 7
'he University of Moscow claimed in7 that he was minibrus-2
for hit
c
ne* cornout"expected in
InS annxdoomed by
MCfetU thatnlikdy tonto ptoductioo aatil tbe.ntended to bare ap lotocMioti.peedillion operations per tecood--fl capability comparable to the US Cray-1.
Soviet performance plans for thereexaggerated. Tbe SorieU have greatlyperformance capabilities for all ETbrus computers. The Soviets have announced theoretical computation
ngingillion "operations" pet sec-ond for the standardp toillionkx-.s" per secondultiprocessoritlioQer second for the futureur analysis of the specifications for the El'brui reveal) (hat these "operations" arc far short of useful, full-precision, floating point arithmeticThe ETbrut I.ingle rxoocssor, achieve*
onlyFLOPS for typical benchmarkcomputing problems. With (wo processors, we
computearginal improvement toinn's. Evenprvxessor configuration, weenchmark ofFLOPS for (be
n the basis of Soviet rxrfotmsnce claims
enfold speed Increase relative to (be El'bnu-l.
wcpeed ofMFLOPS for the EI-brus-2
Soviet research continues to concentrate on achieving high computing speeds through the use of parallel architectures, accordingide varieiyC
Jopen-source articles. This approach allows the Soviets to shift the design burden from one of their wcahncties design and production of sophisticated electronicone of theirmathematics and algorithm design. Instead of using densely packaged, high-speed VLSI logic and mcmeey chips to achieve "supercomputing" speeds, tbe Soviets will attempt to interconnect multiple numbers of relatively slow, unsophisticatedThe problem of achieving high speed, essentially dividing the work evenly among the individualis left to tbe designers of the operating system and application software
Although the Soviei Ei'brus computersmall number of large processors, new systems under devtlopaseat will link tens to hundreds of relatively small microprocessor! together. One of tbe mote promising lines for Soviet achievement or widespread, highspeed computing is MARS (the Soviet acronym for Modular. Asynchronous. Espandib'cARS is an attempt to link together up: bit micro processors This attempt parallels researchbeing done in the West and Is Intended io satisfy
tbe renuiiemenuroad range of Soviet scleniistc and engineerscu ire high-speed computing. MARS' general characteristics indvsde:
lligh-ipeed scientificaps blli lies.
Flexible architecture for applications with different computing requirements.
Object-oriented approach to dataacilitate and automate programming and to provide (be basis for such artificial intelligence (Al) components as knowledge bases and icference engines.
Highly user-friendly iaierfaocs lo maketerr. easy to use byorkers.
The heart of MARS willbit Kjocoswhich (he Soviets have ortodeled after the British-midc INMOS transputer. The transputer and Kronen are essentially microprocerssors withfeatures, primarily communication buses, thai(hem to be readily buili into networks and arrays for parallel pro ceasing..
a single-chip KronoserasUooly in prototype. Dcvelop-menl work on MARS Is continuing by implementing (he Kronen design with multiple chips. Plannedbetween tbe Soviet Kronen and the British transputer has been demonstrated by substitution of (be Iraniputer for Kronos during development work on MARS
White emphasizing development of muttiprocexsoe computers, there is evidence of at least one effort in Ihe Sonet Bloc toray-like supereoertpuier.
r
project to
ray software-compatible supercomputer was Initialed sometime5 and involved (he Soviei Institute of Information Transmission(1PPI) and (be CDR Central Institute forand Information, P
11his project was stalled so that iLs Sen iei Bloc could take advantage of Western softwarewritten for (he Cray. Tbe head of (be project claimed (hat IPPI had acquired aome Crayt; had also acquired Faitxbild ICs used la the Cray and wai reverse-engineering them. In cooperation wiili ZKI
The Soviets have developedvariety of arrayfor connection to their mainframe andAlthough lacking the overall Heaibility of general purpose, high-speed computers, arraycan offer significant performance bene6is in certain applications, such as signal and imageand in the processing of seismic data. Published per for ma nee parameters indicate that Soviet array processors are eight to II years behind the West in overall capabtlittcs. Even if Soviet Bloc array proces-sors are able to achieve claimed capabilities, which we doubt. Soviet ability to proems data, such as that from radar, is. by Western standards, limited. In addition, perhaps an even greater Soviet problem is the size and weight, rather than the speed, of their arrayThe capabilities of Soviet machines that take up
several racks of space can be achieved in the Westew plug-in boards (sec. This offers enormous advantages when tbe system Is to be usedilitary platform for real-time processing of data
Several other high-speed computer! have beenin Soviet open literature, although there is little srocific information on them. Theultiprocessor jointly developed with the Bulgarians, was citedSS Soviet and Bulgarian scientific journals aserformance ofLO PS. The, another multiprocessor system
Figure 22
ad In Magnetic Disk Technology
jcoo
vohc/bioc
ProSutllon or compmble
s
OI
I cm
1MB
IBM0 MB
omb MIB
ProdKlrOn
Jrl
IBM Mil
us HI'
HW ibmjjh
ibmijio-iimb
MO Ml
Ka mb
developed by (be Soviets and Bulgarian!,ofrray processed connected to anainframe. According lo Soviet Bloc preaiingles claimed to be capable ofFLOPS Tbeith an announced speedFLOPS, will connect uprocessors and looks Eke an attempt toachine of (he minisupercompo(er class. It is likely lha( mosi of these machines are, al best, laboratory prototypes. We do not know whether tbe Soviets will developin full-scale production for widespread use. In addition, given past Soviet cugs'cralions, we expect actual performance capabilities of these machinesill lar short of anrtounced claims.
Peripherals
Mat wit Phi Drlret. The most serious computer hardware problem facing the USSR and Eastern Europeack of disk technology. This deficiency limits ibe performance of their computer systems in many applicauooi. Including military applications. The Soviets and Bulgarians, the major producers of large disk drives in the Wirtsw Pact, are0ears behind the West in high-performance magnetic disk technology (teeince ihe introduction of ihcircgabyte (MB) driveevelopers have achieved little more (ban evolutionary upgradesechnology nearlyears old in the West In ihe wake of rriovemenu to inttodoce rmcrocomputeTs throughout theirthe Soviet Dloc is facing (he same difficulties in supplying small Winchester (sealed, nonremovable disk packs) drives for PC*
-Secrtf
and maintenance are two critical aieaa of failuic fofisk drive producers. In particular, Bulgarian-made diak drives have become targets of icvctc criticism throughout tbe Soviel Btoc because of thdr unreliability. In tbe3nvea were aupeooe lo Sovielbeow teerns to be true. According
issstiifactionigarian disk drives is earning other Soviet Bloc countries to begin their own production. The Bast Ocrmans in particular have emphasized development ofproduction facilities.
Bulgarian drives have several major design Saws:
Poor servosyttera controls and system enclosures.
Inadequate filtering systems and startup procedure controls.
Faulty spindle brake systems.
A technical evaluation { the
B disk driveevealed poor overall quality and crude construction The drive had not been tested before packaging aad delivery aod would not operate once il was uacraled
Production of disk drives is concentrated in ibe USSR and In Bulgaria; other Soviet Bloc own tries have only limited capabilities.MB drives probably tbe only disk drive that is produced In sufficient quantities to satisfy the needsroad ipectrum of users. Complaiats about the availability, reliability, and inter changeability of pacts, however, conlinoe, even for this model. Major productioncenter on ihc disk coatiog technology, the disk head manufacture, and the mechanical positioning systems.
The highest capacity disk drives In volume production arcD units from Bulgariand the USSRith only the Bulgarian unit ia widespread use. la addition to its meager storage capacity, this driveow data transfer rate, limited avallabilitv.igh rate of falhueC.
3 that delivery ofrive may take up to lis months after one it ordered. There may thenurther delay of levcral monthso Install the drive. The Bdgarlan
B drive.is the sizemall washing machine,B drives produced In the West can be held in one hand (sec- Bulgaria has been manufacturing small numbers of tbeB drive using WinchesterB Winchester drive (theas been developed, and limited prototype production may have begunS
Development work Is under way in several countries oninch Wincheater-type drives. Tbecihlbited an MW-IOOOat5 Budapest trade fair that ihey claimed wst the first Winchester
Secrete
Western Dltk Drift Technology at Stara Zagora
disk drive plant at Stara Zagora has been the major supplier of disk drives to the Soviei /Sloe since the, when Bulgaria was assigned this task by CEMA. Bulgaria elalmt lo export half of Stara Zagora's production to ihe Soviei Union, with ihe remainder being used In Bulgaria or soldlit Warsaw Pad countries. Since openingke plant kai continuously acquired Weslern technology because of the almost total lack ofBulgarian experience In high-technologyAlthough this Influx of Western technology has allowed Ihe Bulgarians to maintain theirposition In the Soviei Bloc, mistakes Inof the Imported technologies and substandard Indigenous engineering support haveS-year lag with the West
f*3
The most recent series of acquisitions ofsupported modernisation of thebeganhe project wasto be completedS. but delaysexpansion pushed ihe completion date
C Jot uatTiOWestern firms.
supplied technology, much-
which was COCOhf-controlled. to the Stara Zagora project. Western help spanned ihe full spectrum of manufacturing technologies needed for the'indfinal assembly of hard disk drives. Among
the key elements provided were engineer Im assistance 'plant design and planning, training In operation of ihe plant, and design and construction of cleanroduction aisis lance (computer conirol rooms for disk drive and IC production, andmanufacturingnd manufacturing equipment (disk coating equipment, bonders,cleaners, abrasive and polishing machines, grinding machines, milling machines, and lestfor finished disks)
Although mosl of Ike upgrading at Stara Zagorathe production of large, relativelydrives "
- - ls facility Is designed to produce up0 hardonth, giving the Soviet Bloc the capability lo mass produce small Winchester drives If they can alio manufacture the read-write heads and mechanical components that the drives-require. The small size, large storage capacity, and high reliability of Winchester drives Is particularly Important for use In PCs and In military computer systems. We expect these technology acquisitions to be reflected In Improved capabilities of virtually any Soviet weapon system thatomputer.
developedocialist country. The drivelaimed capacity ofBs. The unit shown was describedevelopmental model,rototypemall test series scheduled to be builteries production was expectedut we have no evidence (hat this has occurred
The East Germans daim to have dcvckuxddrives for use with Warsaw PactThree different models withties rant ins8I MBs haveWe have rv> -videnoe tltit these drivesscries
C' many was to begininchester driveS-MB capadty for PCs. tbe8
Although the low performance of Bulgarian and Soviet disk drives may Impose only inconveniences for the majority of computers in use, system performance of current and future Soviet high-speed scientific and large general purposeeing and will be severely hampered without further advances In disk technology. Some military systems requiring the high-speed
processing of large amounu of dau most likely have been negated, delayed, ot reduced in capability be-cause of dbk drive deficiencies. Theost leiidui on large, real-time computational applications, such as ballistic missile defense, and on high-volume, hitb'tpeed dau transfer applications teqtiri.it large data bases, aucfc as command, control, aad cornmtini-c* lions systems.
Happy Disks. Despite concerted efforts toof floppy dbk drives and disks, botha rare commodity In the Bloc. FloppyIn the Soviet Bloc lags that In the Westfive years. Many floppies producedusing single-sided, double-density techniques.format dbks still widely used in the Sovietbeen made obsolete In the Westneb floppies In addition toof availability. Soviet Bloc floppy disksdrives have very poor reliability, liowerer.of several turnkey floppy disk drivethe West will likely irnprovr, availabilityin the Soviet Blocfacilities available to many civilian Soviet aad East European computer users. Although theragmentary for miliury installations, we believe that upe drivesimilar function for many of these users
ittle technical or other cpuotifiablc dau available on Ihc prcductioo and use of magnetic computer upe In the USSR and Eastern Europe. Tbe USSR and East Germany produce magnetic tape, reportedly using Western wchnology and equipment. We do not know the production rates. Bast Germany advertises production of computer upe capablepi recording density, although wo have not been able to subsuatiatc these claims. Complaints of poor quality aad iaadequate supplies of indigenous Upe persist; users uwforoily prefer Western Upe, which we believe to be widely available
Targetiag iht Weal for Advanced Hardware
Matmiie Tape Driver. The slate of the art iatape drives in (be USSR andropeits per inch (bps) densityau transfer raleBs per second Comparable equipment first appeared in the United Suteshe nextra5 MBs per second, has been in use in the Westhen the problems of drive quality and reliability and of upe quality and availability arc considered, the cornpari-son would be even more unfavorable to the Soviet Bloc. Although modelseniltypi have been announced in Bulgaria and have recently been rumored to be available In the (ISSR. there is only limited evidence of (heir use
Tape drives are the external storage devices moat eomrrjooly used with Sonet and East Eauopeansystems. Given the problems "Iih dbk drives described earlier. Upe drives arc often Ihe only mass
high-;
Bign-pfiorily program to acquire Westerntheir technology has been in effect since theDuring thendmphasis was oa IBM mainframeacqubUions. mainly iJkgal. led (oand production of thend Ryadof IBM-eompatibleimilarly,acquisition Uriels daring much ofmodels, critical to Sovietof DEC-ocenpatibk SM nunicomputcrs.Soviet emphasis has been on acquiring.lupcrmuucocipciter. andtechnology. Although US computerbeen the primary Soviel target, otheralso been urgeled.
Many of (bete countries produce hardware orcompatible with IBM and DEC machine*in other programs that are of Interest to
he primary computer
target for Soviel illegal acquisition inai been (he US VAX family of superminicomputers (tee. The St/viets use illegally acquired VAX computers directly in military-re!Jtto activities such as CAD/CAM for integrated circuits, mechanical design, and aatotnaied manufacturing VAX super-rrurueoeaputert ore also used ia Soviet enmmunicu-tions and image-processing activities
The Swedish and German seizures in late IMS of VAX hardware, software, and related equipment desiined for Ibc USSRrime ciamplc of Soviet attemptscquire computer technology.
3 Vest German and Swedish authoritiesargo ship at the Hamburg and
H citing borg ports, respectively, and removed various computer equipment. The equipment was originally shipped to West Germany from where it was illegally shipped on to South Africa for testing andbefore proceeding to the USSR. Tbe equipment seised was valued at nearly SO3 dollars and incloded seven Urge VAX II/TSO- mega byte disk drives. TO ccoor graphic terminals, artwork and either hardware, aad software. On tbe basisetailed study of tbe hardware and ooftwarc configurations, we believe advanced semiconductor research and development was the intended application
illegal acquisition of* VAX computers ind legal and Illegal acquisition of other core pater technology hare contributed wbstantialty to Soviet computer caps bili lies:
have allowed the Soviellinimize Ihetr lag andodern dau-proccaslng capability. By acquiring and crapying Western tech-nologj. the Soviets have incorporatedalbeitbauatia) lag. id theirsystems.
copies of US cornpulcr systems have been widely used by the military
Illegally acquired Western systems have been used directly in high-priority military-relatedinch as computer-aided-design (CAD) for weapons research and design prograrrn "
Soviet Doc Needs
The Soviets will need toide variety of indigenous technologies to match Western computer develop meals (see figure 2JI Soviet gains In central processor unit (CPU) speed, which will come primarily from denser packaging of higher speed circuits, win require Soviet advances in ICs and nunefaaaiirtg technology. Tbey will need to develop atoderate level of VLSI production capabilities and expand (belt production capacity for all types of ICs. In addition, the Soviets will need substantialin their capabilities for designing andmultilayer printed circuit boards (PC Flu As the packaging of circuits increases io sophist re* lion, the manufacturing process also becomes more complei. This will require greater attention to quality control at each manufacturing slep and increased letting during assembly
Soviei problems in semkooductor electronics areby continuing illegal acquisition of upillion relatlve'y low-technologyear from the West, suchit microprocesjcrsBKBther dements necessary for mass production of microcomputers, particularly precision production technology and automated production and test equipment, are also UcUng la ihe USSR and continue to be priority targets for acquisition from the Weal
Soviet Bloc countries will need to improve disk drivecapabilities to lake advantage of any Increase in computer performance. Increasing (be data transfer rales and storage en parity of Soviet Bloc disk drives will require more sophisticated read-writeisks that are machined to higher tolerances, improved costing formulations and applicationfor Ihe magnetic medio, and greatly improved Quality conirol.
Deiii* rVrAao/egy. Wc bdieve Soviet engineer* will coalinue to depend on Western computer* as their prime source of design direction. Reverse-engineering, however, will become increasingly difficult, requiring eslensive adaptation and comprorniso. In addition, the Soviet* will need state-of-the-art CAD hardware and software from the West, Indudingt
CAD
Application software for
Logic design and verification.
System simulation.
Wiring layout optimlutlon-
Prototype design
Clients. The Soviet computer industry is constrained, in part, by gross deficiencies In (he scniiconductor industry. The Soviets need lo increase the quality, reliability, and quantities of ICs tbey produce. The goal of producing millions of PCs by tbeill require increased Soviet production of truer uceoeeajcrs and memory chips. Advances in rnaiafraiDC* and dcvelopcocot of supercomputers will require fsster, more reliable ICs and greatly expanded VLSI capabilities to achieve required drcull density. To meet these goals Ihe Soviet* will be targeting the West for:
Automated crystal puller*
Clean room technology.
Utrtograohy and etching equipment.
Epitaxial and orsnepitaxial deposition equipment.Automated tesi equipment
Multilayer Printed Circuit Boards. Matchingcomputer advance* will require significant Soviet advances in packaging electronic components-Improvements will be needed in Soviet ability to produce mrdUlayered PCBs and to design andcooling techniques to dissipate the beat of densely packed ICa. The Soviets will be looking to the West for help In increasing tbe sophistication of PCB production and in increasing production capacity to alleviate an overall shortage. Surface mountfor ICs will also eventually be targeted by the Soviets to achieve needed circuit density required In modern computers. Targeted Western technology will include:
CAD software for design of multilayer boards.
Digitally controlled multispindle drills.
Equipment and technology for dry methods of PCB production.
Equipment for testing multilayer boards.
Technical know-how for dissipating beat incomputers.
Technology for surface mounting of components.
Automated Manufacturing Technaiety. We believe that the. Soviets wQl seek to increase the level of automation in manufacturing processes (sec figures. Many of the Soviets' computer production problemsirect result of poor workmanship during the manufacturing process. Automation will increase productivity ond product reliability.it will facilitate production of more advanced computers. Automated production equipment and technical know-how will be key targets in the West, which specifically Include:
Improved plant management techniques.
Information on inspection and test procedures.
Automated Inventory controlnformation on quality control.
Equipment for automatic insertion of components.
Automatic wire harness fabrication machines.
Automatic soldering and wire wrapping machines.
Automatic test equipment for
Circuit boards.
Subassemblies.
Backplane wiring.
We believe Joint ventures between the Soviets and Western firms offer the potential for providing much-of the manufacturing know-how the Soviets need for improving computer production
asset I
Source* of Supply
Over (hebe Soviets will look to ihe United Stales. Japan, and Western Europe for the vail majority of (heir technology accds. However, newly industrial tied countries, especially (hose in Asia, will become increasingly attractive targeti to tbe Sovicti for high leclUkoiogy. Although COCOM member countries maintain aubstantial leads la comovercial and military computer technology, non-COCOM countries are eapanding the supply of commercial computer technology. The Far East alreadyajor supplier of PCs to the Bloc.'
The United Sutes, Japan, and Western Europe now and for the forseeabie future hotd the lead In the underlying technologies that drive computer production:
Serrucoavductor device and component technologies.
Interconnect technologies.
Mass Storage technologies.
Network and communication technologies
However, many of the newly industrialired countries are Investing heavily In these and other high-tcchnol-ogy Industries. Even without sute-of-the-artthese countries will become increasinglyUrgcUhe Soviets a* Western countries tighten eiport control*
The production of PCs and their peripherals la the newly industrialircd countries is particularly notable Countries such as South Korea. Taiwan, andin tbe Far East, as well as India and BrarJl. are close to the forefront of PC technologies and can rapidly mimic Western development! In many casei this Isesult of legal transfer of (he underlying (echnolofW by ibe United Suies and other COCOM nations
At the other end of the computer scale, tbeof bigh-speed computers worldwide will Increase Soviet proapecu of reducing the Western lead in this area. Although the United Sutes and Japan arc Ihe only sources of supercomputers, manufsctuien in Western Europe will probably enter the market by the. New manufacturers la the United Sutes are offering wider ranges of prices and capabilities.
Some machines now otTerraction of ihc cost of traditional rtuchinea. and future developcuents will lead to even lower costs and higher performance '
Sovieto AcnauSJIIooi
The SovieU' massive, writ-organized campaign for acquiring We*tern compute* technology consists of two separate progratns, one managed by the Miliury Industrial Commission (VPK) and one by Ihe Ministry of Foreign Trade. The VPK program, primarilythrough intelligence channels, seeks computer hardware, blueprints, product umptes, and lestthat will help Soviet designers rtvtsse-essgiaeer Wester* computers. Figureypical VPK requirement that Urgcted technical clocu meet* lion on tbeystem, information that was eventually used in the design of Soviet Ryad II nulnframes.
The Ministry of Foreign Traderade diversion programcquire manufacturing and test equipment to Improve production capabilities. This program depends largely on Western traders tothe Soviets with Western technology, ranging from individual computers lo entire turnkey prodoc-tioe line* for terokotsdoctora, disk drives, or printed circuit boards We believe Soviet computer advances will depend heavily on the Soviets" ability to acquire and use Western production technology
When possible, the SovieU will seek tombargoed Western computeraod software, particularly for CAD/CAM.support the Soviet Bloc computer industryresearch laboratory and on tbeOcrsnanv snake*
extensive use of DEC VAX computers aad Tekironii vrorUutioas In their reverse-engineering of Western hardware. In addition. Western computer hardware will continue to give the Soviets' research anddirection and serveenchmark for Soviet measurement of their Indigenous computer develop-menls
-Srvm
f lo ihc 'traditional" methods of technologyuncul bat created new opportunities for ihe Soviets. For instastce.ventirea with Weatero firmsotential avenue ford: variety of Western technology. Even joint ventures for production of relatively unsophisticated computers will provide the Soviets with management andknow-how that will hrncrtt tbe manufacture of morecoss
Although many joint ventures arc being dlscuaicd.ew minor agreements have been signed, according to Intelligence and open press reporting. Westero businesses appear to be encountering tbe traditional problems in negotiating with the Soviets, including access to Soviet markets, convertibility of ruble profits, legal jurisdiction, and other operational
matters. The ultimate value of joint ventures to the Soviets will depend on how successful they are io working out these problems
Mot ventures offer tbe potential foe not onlythe Soviets with Western productionalso giveestern partner who coulda conduit for Western production equipment
virtually every joint venture calls for initial production to be dependent on imported Western components For eaaninle. in8 Ihe Soviet Ministry of Instrument Making, Automation Equipment, and Control Systems (Minpribot) wasestern partner to cooperate in designing and manufacturing
i ikw .erics
ihe Sovicu wioi u>
develop;
A design center Tor PC component* and raicrocke-tronic device*
CAD scrftware to lonporl product design.
A printed circuit board fabrication facility
A printer assembly facility.
A hard disk drive tmnufacturing fadllty.
An automated PC assembly facility0ear.
These developments will supportuf*ctureew. three-modd series of PCs. The primaryo be configuredAD workstation and will be IBMurther requirement is support for parallel development of more advanced products to prevent future reliance on obsolete technology.
The Soviets will also have access to an ever expanding market of used high-techrK4ogy equipment formany of ihdr Western technology requirements. This equipment ranges from computer hardware to pteduction equipment. At an example, much of (lie Western production equipment supplied to thefor hard dish production at Sura Zagora ni secondhand
There is great incentive for the Soviets toWestern technology, particularly computers.the computers arc considered "outdated**standards, in reality tbey are veryearsihe used Westernyears more advsrsccd than what theproduce indigenously. Prices forraction of thai for newinstance, usedhat listed newsell used for as little asnis easier to divert lecondharid computers,(hose originally sold new to European endUS reexport
licensing requirements are virtually irisorcdo-pean secondhand computer sellers |_
llege that falsification of cod users in export licenses is so common thatestination in Austriaitter la adsed computer is almost assumed by them toiversion. *
Outlook for CMtlaa, and Military Programs
On the basis of past pcrfcross ncc and of out current assessments of Sovietu?abiUiies, we expect tbe Soviets to fall further behind the West in computing capabilities throughout, LC our view, the Soviet bureaucracy will find ii difficult, despite Gorbachev'* modem Italic* program, to sake (he necessary step* to rxxrect snaay of tbe computer industry's wcU-rccogniied problems. Therefore, we expect Soviei reliance on US and Western crxmptiter technology to persist and illegal acquisition attempts lo continue. If the Soviets obtain larakey production facilities or detailed production know-bow from ibethey have done In (bemay bo able to narrow, al leastpecificgap
Recent organizational changes and innovationsto address scene of the problems of the industry have been Largely umucceasf ul.6 ibe State Committee for Computer Technology and'OKNTT) was est*Wished to foster coordinated computer productioneplications Tbe Depart-rneot of Informatics. Computer Technology, and Automation of tbe USSR Academy of Sciences will attempt to focus and concentrate fundamentalin information technologies. Also,ll-Unioo Scientific Complexes (MNTKs) were created lo speed up Ihe process of Innovation in (he Soviet economy.IK for PCs it charged with ovga-niring and coordinating the development andof PCseet (he (oral requirements of the national economy. Success demands the nearlytask of coordinating the efforts ofifferentthat develop and produce compaters andthers that make materials, parts, and cempo-neots for then)
We bdieve that ia tbe future the Soviets will be forced to incorporate more advanced technology into their weapon systems to stay competitive withmilitary dneropmentS. Wi(houl significantin computer technology, (he performance and reliability of these future Soviet weapons systems
- -'
could be jeopardised. For example. Moscow's ability to develop c* acquire advanced computer technology willirect Impact onbility to deploy improved antiballLiilc mlulle ratlaii for reentrydiscrimination and multiple Uriel handling.capability to design and produce compositefor the next generation airborne aad space vehicles depends indirectly on modern computers ia tbeand on the plant Boor 1
Tbe weapoo systems moat directly affected will be those that require fast com paters with huge mala memory, fast dbk drives, aod tbe ability lo eachange dau with outer computers tUamples include cocrt-mxui, control, crjcnrauokatlotu, and mielligeaccn oa the front and theater level, mobile command aad control ayttemt, intelligence andsystems, and airborne warning and control system (AWACSL Other weapon systems such as Unks. planes, rrussllea. and tactical communications are dependent on computer*ess direct way. These systems Incorporate microprocessors, memorynd other IC* that generally require computers toem.
Engineering worksUtioas are needed to Implement computer-aidedas CAD, CAM, aad computer-aidedare key elements in modernisation of tbe Soviet miliury and civilian aviation industry. Soviel managers and miliuryhave beoxpe increasingly conversant with Ihc role thai CAD can have en iiraproving ihe availability, reliability, and performance of weapon ay stems, while reducing tbe human and material resource
Implementing Increased CAD usage will requite not only engineering woAsUtioos butomputer-literate work force that can take advanUge of them. Without large numbers of cotnputers for education, the Soviets cannot implement their puns for ex tensive training programs to prepare engineers, designers, and technicians working hi CAD, CAM. and CAE (computer-aideds well at computerirrd management information systems. These programs will require significant development efforts and lub-suatial improvements In the quantity and quality of production of Soviet computer equipment
Appendix
Tables of Computer Parameters
USSR: Sekclrf MJcroc^iwfen
II
toll
If
12
ES-lt4|
| 23
16
2)
>:0
J
-
Ml
to 64
m :;
hi J
fl)
fall
MHfrtQtO*
?7H
iqiw
j'j lit
marsjsmq)
hotolo*
510
mo*
110
SM-OI
I0IO
is
pro;
moo*
ton-
pc
tOBO*
as
lmhmk-it
Table 4
Hungry: SdecW
Moid
Model
100
16
X
11
Z*Q
MW4-W
ftm*
-
2
rim-i'ui
PC/AT
MtftOOO
16
tiid5
1 1
-
SP
to 64
1J
CPP
16
and MotOCO
16
64
PC/AT
*
K4
A6
PC
64
3c
IniSL-d:
Ifiuftuy: Selected Microliters
MoJcl
Uofio -
With
USlvteJO^
'
400
700
Ctattr
*
16
'-'O- < j
-
Prcfi-S
_
I6M
+ t6
Mirto
PC
XT
IV/vr
*
16
Table 5
PoUmt Sckcftd Microcomputers
With
US ModrJ
XT
UlfO
'
PC/XT
PC
j 4H
liMIX-t,
* HMO
fv
IDM PC
^
Id
KOtEKaKT
."dim?
i
PC
pc/xt
toil
PC/XT
irvl IOU
MCY
tlrs*0<
PC/XT
1 uit German
* Potiih;othcri Arc US.
u ti
TiHftC
Selected Mtaoctraptittn
Model
Unf Hi
Model
64
iooo
PC
to 44
PC/xt
1oi.0j4
-
PC
i03t
PC/xt
im
234
PC
16A
*
PC
tCO
16
-
PC
!6C
PC
16mm
wt son
PC
650
ll
pob^hed: tfchfilallr,it micrcpeoccuor capu dl rootyitobylaof
<So*irt
Gerra*n; all otWa are US.
USSR/EuCtra Europo; MlricompuienSup<rainkomputm
Model
Orlivcrod
Umttj
2IXX
JIXX
CietfrotJo^Ui. Poland, Rcaab
in 5C
Ctotfeoilovatia,
.
J
Family
to 4
11
loS
16
mi
4
USSR
16
0 to2
SM50O
Cxotfc*krrikii
t
Germany
tot
4
oo i
o
236
to
toil
Hunncv
in
H
LfcHrpnlfci-|2
fcoaOentufly <jssr
J
3jtooo
0
Table 9
USSR/Eutern Europe Rrido4nf*(en
IQH
4
I la i
MI
0 5
mi
1CK>
ten
JWJ3G
J*C/
7
11
MO/40
36C/
)i
11
o$6
L'ssr
niind
USSR
I
ft/td II
/ MJ71
:
MM t
Q55M
es
ussr
KM! LL
itejtyis
WMHWIS3
M
>7C*
; M4
mi in*
*f!
n <
io*;
2 id It
111
ma
111
0}?
* 1
ttny motor
libit 10
USSR:doiitifc Computer,
rw^voy
Drb.tr oJ
nets*
a 1
1
7
TibUll
CSSR/Eoittni Eomtc Amy Procraort
w?
It451
IUS4J
Hunprj
USSR
ES-igw
S
ES-1QJS
oitjif poim opemJotu per
Tmbk 12
USSR/Eastern Ewc-pc; RigidUIMik Drires
Wood
t
Kit*
UffiiiijgbMM*
:osi
OC
Otijoj
USSR
IS
IM
75
it 01
IV)
FA-
not
prctol
300
106
i3
1
WO
^4nej tl
Table IS
USSR/Eaneni frrope Floppy Disk Drift*
Model
bpi*)
SCS5
VS-
e ^'
a hi mi mi a m
,
SS/DO
Tnti* per^Hatted,
bin perDouWc deAfJiy.
single ckftsrty.
T.ble 14
USSR/Eastern Europe; Magnetic Disk Packs
of Ntimberof Trtdu
mvrw-^rro
-
20Q
t
Table 15
USSR/Eaateni Europe: ReeMa-Reel Magnetic Tape Drives
Original document.
Comment about this article, ask questions, or add new information about this topic: