Sower"
RELEASE AS SANITIZED . 9
Soviet Progress in the Production of Integrated Circuits
-Seeret-
s
SOVIET PROGRESS IN THE PRODUCTION OF INTEGRATED CIRCUITS
SUMMARY
The USSR is running hard, but so lar futitcly, to catch up with the United States in the field of integrated circuitsroduction is less thanf US output, and the USSRoears behind the United States in ICtechnology.
Although Moscow has vigorously pushed the development of ICs sincearge-scaleillion or more devices per year) of simple monolithic ICs began onlydvanced types are at bestonlyilot basis.arge share of Soviet annual ICas much asnot meet design requirements or quality standards.
Production yields are low because Soviet plants
lack advanced manufacturing and testing machinery and techniques,
maintain poor environmental standards and quality control procedures, and
employ backward management techniques.
In most cases, production line equipment is ol domestic origin. The most modem and productive Soviet IC production facility, the Mikron Plant In Zelenograd, is equipped mainly with Western machinery acquired illicitly outside embargo channels, or from nOn-COCOM countries.
The Soviet armed forces control the production of integrated circuits. An overwhelming share of ICs goes into military/space equipment such ascircuitry of air defense missiles, avionics equipment on fighter aircraft, and computers with special military applications. Because long lead times are needed to redesign military electronics hardware for ICs and to interface new-gencration electronics wilh existing military systems, ICs are used only selectively In these applications. No new generation of military electronics equipment basedon ICs has been identified. The brand new Soviet military hardware captured during the recent Mideastmissiles, radios, and ground communicationswould be logical candidates for mi-crominiaturization did not incorporate ICs.
ICs have been sparingly used in Soviet civilian equipment. Civilianfor third-generation computers, which potentially would needillion units per year, are just getting off the ground. At most, no moreillion ICs were needed2 lo meet all civilian requirements.
No.t:iclha publicationy puythe Office of Economic IWaich.
G. The USSR, probably hoping lo exploit East European acquisitions of Western IC production know-how. hat sough! joint IC development andagreements with East European countries. An agreement with Poland, lor example, followed quickly in ihe wake of Warsaw's successful negotiations with Franceurnkey IC production plant.
The USSR for many years has pursued every available means of obtaining IC production equipment and know-how from the Wat. The reliance on Western equipment acquired through various channels helps to account for both Iheand the failures of the Soviet IC program. Western equipment has filled gaps In the production process, making IC production possible at an earlier date. The piecemeal acquisition of technology, however, without supportingknow-how probably has reduced yields and the quality of the devices being produced. Substantial growth in the future in the quality and volume of IC output hinges upon the acquisition of Western know-how and equipment. The Soviets,onsequence, will continue their intensive efforts to acquire Western equipment and technology for the manufacture of integrated circuits.
Unless there are substantial reductions in the international embargo of semiconductor production technology and equipment, Soviet progress inarge and technologically advanced semiconductor industry will remain slow. Despite occasional diversions, the embargo has been generally effective in denying the USSR access to the advanced production technology needed to overcome its deficiencies.
DISCUSSION
Introduction
The explosion in integrated circuit technology over the lastears has revolutionized the electronic industries of the West. This revolution has made it possible to produce equipment of ever-incrcasing com plenty and with size, weight, power consumption, and costs sufficiently small for practical useide variety of applications. In particular, this new technology has permitted the development of military weapons and support systems not previously feasible.
Since integrated circuits were first developed in the United Stateshe USSR has tried to develop an indigenous technology and to take ad-vantage of the West's superior capabilities. This publication surveys the integrated circuit industry of the USSR in terms of the quantity and value of production, the types and quality of output, and the patterns of consumption. The publication does not discuss scientific issues in any detail and should not be taken as anof Soviet progress in integrated circuit research and developmentrief discussion of how Soviet IC production was estimated for recent years, andontains basic terminology for the nontechnical reader.
Background
United States is the worlds major producer of integratedthe world leader in integrated circuit engineering andS firms, including wholly-owned foreign subsidiaries,circuit devices valued4 billion, about throe-fourths ofor four limes the output of Japan and.imes the output of Western
United Stains
Europe
YfiefercOnp owpur or (evergn juWWe.rci or USrape wotlae-
Europe (see1 Output of ICs in the USSRaluer! at the average work! market pitce. amounted to SOS million to Wj million, leu thanf US output.
VI. The United Stales also is (he major reporter ol ICs.Cs valued at0 million were exported, mostlyestern Europe and Japan. US exports and sales abroad by wholly -owned US subsidiaries supplyf West European and one-third of Japanese requirements for ICi. These shares may be increasing. US eiports ol ICs during the first hall3 were up
billionttriiinviiduetot output.
dramatically ova ihe corresponding periodalesestern Europe doubled and ihosc to Japan were up by three-fourths. Increased sales to Japan, stimulated by the easing of Japanese import restrictions, could further strengthen US domination of the world Integrated circuit market.
he United Slates has pioneered nearly every major advance In IC manufacturing technology. Through sales of patents and licenses and the direct transfer of know-how to wholly-owned subsidiaries. US manufacturinghas spread to Western Europe and Japan. Nevertheless, because ofrapid advances in semiconductor teclinology in the United States, no West European country or Japan has been able to achieve overall technological parity with the United States.
Although integrated circuits were initially developed for militarythe share of IC output used for military/space appucations has declined steadily in the United States. Rising yields and falling costs bave resulted in dramatic reductions in priceemarkable expansion in cooimcrialThe average price of an integrated circuit fell5uring this period the military/space share of US output fell.
The decline in the military/space share of IC output masks themilitary importance of IC technology. As the reliability and performance of all lypes of ICs have improved, applications of ICs to military electronics systems have expanded. In the United States, military procurement of ICs grew from1 to43 million5 million) while the procurement of military electronics hardware declined9 billion2 billion).
The Soviet IC Program
The development of integrated circuits in the USSR has been pushedriority basis sincelthough interested in IC technology as earlyhe USSR lacked the precision equipment and technical know-how needed to translate known scientilic principles into laboratory prototypes. Soviet microelectronics technology was still rudimentarynimical handbooks and other literature to aid designers were unavailable, and specialized facilities for training technicians and designers were practically nonexistent. Also, having opted early for germanium as the basic raw material for the production of serni-condiicioi devices, the USSR failed to exploit the llceD experience in silicon epitaxial and planar processes that had been thr bam for major semiconductor developments in nOn-COmmunist countries.
In theecision to push the development of silicon technology may have been made in thehen cmuhuction beganarge new development/production complex lot the USSR Ministry of the Electronics Industry at Zelcnograd. This complex, which was builtriority pace and came into operation. has become the leading scientific center in the USSR for the advancement of silicon semiconductor state-of-the-art. Zelcnograd produced the first Soviet silicon epitaxial-planar) and theonolithic IC based onubsequently it has carried out major development work on hybrid ICs, bipolar transistor-Uansistor logic (TTL)
devices, and. most recently, metal-oxide semiconductors (MOS1 and emitter coupled logic (ECL) devices.1
The USSR has continued vigorous cflorts to develop monolithic ICs Iromsing planar epitaxial processes. In the United States, similar efforts were abandoned by mostea.ly inart, the USSR has succeeded where US firms earlier had failed. Germanium ICsdeveloped in thend apparendy now are being produced in small quantities Yields, however, are reported to be very low and quality and reliabilityThe highly efficient production techniques lhat work so well with silicon are nol readily adaptable to germanium.
8 the Soviet IC program, apparently with high-level backing and pnonty funding, was in higharge number of Institutes and design bureaus had been drawn into the program,ew new experimental plants had been bu.lt to speed the transition from laboratory prototype to seriesSome hybrid ICs already were being produced, although in only small quantities. At the same time, morerototypes of monobthic ICs had been developed. Moreover, epltaxial-plana. transistors had entered seriesindicating that the USSR had assimilated the basic processes for the production of monolithic ICs.
Military versus Civilian Requirements
The Soviet IC program has strong military overtones. One source has said that the great majority of all the ICs produced in the USSR were intended lor military applications. According to this saneenior Sovietesponsible for IC production has stated that "ICs are used almost exclusively for military and space equipment."
Most Soviet design bureaus and plants engaged in IC research and de-vxlopmcnt have dual subordination: to rhc Genual Scientific Institute No.f the Ministry of Defense, which approve and monitors defense-oriented IC development projects; and to the Management (OueVtsiya) ot the Scientific Center.hich plans and finances IC research and development Ihe activities of the Scientific Center, in turn, are at least partially controlled by the Military-Industrialhich, according to the source, -initiates monitors. jukI terminates all defense, dated IC development and production programs' In addition, the military repiescrilalivcs assigned to practically all development and production facilities apparently have broad
' Foi definitions ol <echn*al lenm, ur*
buihl0le,
A miw< ewewion. IBMfi7.ii, geinuiuuni )Ci unlit at leanpbnuit* productionIo.
Cuboid mate lo tit* Mfcl'
minuleni ol defciue.rtbwd .nim.Km.i, ofhci.l, and lechnictlliont mduitiy.
to interfere, if necessary, in the production process. For example, the transistor production line at the Svetlana Plant in Leningrad reportedly was "frequently shut down by military representatives for not meeting specifications."
Because of the extraordinary secrecy surrounding the development,and use of ICs in the USSR, information on major military weapons systems that use or are intended to use ICs is scarce. ICs reportedly arefor use in naval avionics equipment, avionics for military fighter aircraft, navigation instruments for various military systems, and special applicationcomputers. Also, it has been reported that monolithic circuits, nicknamednd developed, are being used in the guidanceof missiles used in the Moscow air defense system. Other reports indicate that ICs. including rnonolithics, have been used toariety of laboratory test and measuring instruments, probably as unique items orind basis for use in the development and production of electronic hardware for military/space purposes.
In general, ICs seem to be used only in the highest priority military applications. Many new items of Soviet militaryradars, and communication?captured during the recent Mideast War and now are under technical exploitation In the United States. The electronic circuitry of this equipment consists mainly of vacuum tubes and germanium transistors (the component technology and design is of theot ICs.
The USSRmall portion of its ICs in civilian electronicsMajor civilian users include Soviet producers of third-generationespecially RYAD general-purposc data processing computers patterned after theeries, and minicomputers designed for industrial process control applications. Because these programs are far behind schedule, they useew million ICs annually. Theyarge potential demand, how.the order ofillion units annually. The USSR plans to micromini-aiurize industrial instrumentation and civil communications equipment using ICs. So far, this kind of equipment has not been produced serially.
Path of Development
development of more complex integrated circuits and newin the USSR has generally followed the same sequence as inubstantial time lag. Indeed, the Soviet IC developmentmay be primarily oriented to copying Western products andof developing nativeholon, MinisterUSSR Electronicsthatother foreign) IC and solid-state technology be reproduced at thein Zelcnograd. To aid In the assimilation of foreign developments,copiedajor US producer of ICs was introducednd vacuunuzation and climatology conditions described asin the USSR were established. Subsequently. Shokin criticized Mikronslow to copy Western circuits. By copying Western IC developments, the
*The chronolofcy of development ii aitmall-icale integration
(SSI)then linear;mi Luge teak iniegralion
USSIl navesosts and valuable development time. While economical, thi* strategy condemns the USSR to remain considerably behind the West in IC technology.
Statr-of-the-An"
ybrid IC. and monolithic
digital ICs ol SSI complexity, mainly DTU and TTLs. Linear IC. probably are being produced in pilot-scale quanht.es. More advanced digitalECU. TTLs o( MSI complexity. ,ndbeen manufactured as prototypes but are still largely in the development phase (see
oviet state-of-the-art in both hybrids and irwnolrthics, as reflected in technical evaluations of Soviet ICs by US experts, is atears behind US technology Progress has been greatest lor TTLs (SSI) and for ECU. and the Soviet-US technology gap lor these devices has been narrowing. For highdcnsily devices MSI and LSI-progress has been slow, and the gap is widening. MSII devices have already been under development in the USSRears, whereas the development cycle in the United States batedears
o,
inwdade-tmiauto. locicl*MQ JBJrruwor If-BoUar h
ears, respectively. Improvcmcnls in the Soviet state-of-the-art fortypes of ICs are discussed below.
Series productionew simple types of hybrid circuits, consisting of either thin- or thick-film resistor networks with one or more discrete transistors, began in the USSRhe technology embodied in these devices resembles that of US devices of the; they now are beingby more advanced hybrids, thin-film circuits employingnd more recently, by several types using up toonolithic IC chips per circuit. These advanced circuits match the performance, although probably not the reliability, of standard hybrid types available in the United States. Soviet state-of-the-art, however, falls far short of many hybrid devices custom-made in the United States to meet extraordinarily complex design requirements. The United States, for example, manufactures multi-layer thick-film ceramic hybrids that use advanced processing techniques" and achieve high levels ofthe USSR does not.
Advances in the design of Soviet hybrids reflect improvements mainly in thin-film preparation techniques. Deficiencies in Soviet thln-flmpeeling and blistering were noted in devices produced0 orto have been corrected. The preservation of the contents of the encapsulated package from outsidestillroblem.
TTU constitute the bulk of monolithic IC output in the USSR and have improved considerably during the past three years. Samples3 Soviet TTLs with medium power and speed have propagation delaypeed) on Ihe order ofanosecondsompared withi for the same scries of devices manufacturedn addition, current TTL devices reflect major improvements in key areas such as bonding, mask-alignment, gold doping, and passivation of mctalization."
The USSR has developed but apparently has not yet produced mono-tithic ECL devices of exceptionally high performance. Laboratory tests on recent Soviet samples demonstrated performance characteristics equal or superior to most ECL devices now available off-the-shelf in the United States. For example. Che Soviet ECLs operatedpeeds (compared withs for common US types) but are slower than the moil advanced ECLs availableustom-made basis in the United7he reliability of Soviet devices 1ms not been established.
ECL devices of the type that have been tested in the United States piobiibly arc not yet available for commercial use. Wc knowrototype of one model In ihe RYAD computerbuilt with ECLsuch lower state-of-the-art (slower and lessecause theepends upon high-speed ECLs to meet its design capabilities, the more advanced type probably would have been used if they had been available.
"One of llirie leehniquei. Ihe beam laidi advancedirol low yfctib in ICand greatly unprovo reliabilily.
"Time requiredpat* an electrical irnpolw limn one circuit in another. In iheadvanced TTUnltky) have resitted veijpeettta and In*.
" The promt oJ protection the drvit* from tiinlamin.ilion and other damage during crilkid piucnum; .ten* by oiMlubig ihe .ilKaa wnfet1 an impe new ableof titkon iHoiMSr I I
he* MSI/UI rnonoliiruc ICi,earsb M'" ! "* Soviet7cv.ee,
J5Uni"dare
.lower and much larger lhan current US counterparts. Because larger devicesheyewer ecu per wafer of SLmteT
P^ive with Wenem-made devices. Unlessmaye com^^^
ZC Undc''he USSR sinceoviet hnear dev.ee, are still outdated by US standards, beingesignto .be earliest device, made in the Unitedorkear dev,ces may have suffered because of the more urgent reouiro.o, digital devices. In any case, Soviet state-of-the-art tolorSc hnear ICsears behind Soviet digital IC technology.
Production
The Bole of ICt in ihe USSR
emiconductor production in ihe USSR still consist,f conventual Uamistors and diodes. Germanium devicesrge sLre of these, posnbly the largest" Thef,le isvjor shifTnder way.integrated circuits are beginningutmantial share of total output Nonrthelesi. outpu. of
COnveTitional temicon-
as oo. ye. peakedikely to continue to dominate the productleast5 (Table
Tail. I
on. | irmieordue-
TiaiiBMora
3,* 00 17
_ _
Total
c
^conveniional semico.Kluctors is dictated by hij-h
demand.ms of electronic, hardware of .ccent design and manufaCufe both cm! andmislors.and eventsoJvcohonl1urongly resist change and prefer to-design around proved component. Moreover, the redesign ofse ICs canrotracted process laslmc several yean.
of the USSR Thmueh
"ucuiilon otArWiOnKr AppcmiU A
of total IC productiongiven ineemto known civil and military end-use, Ko, example, no more than 3would have been needed2 to meet allillion ICs available for military/space use. But wC have notto identify any military program, or combination of programs, thatfor that number of ICs. In the United States, which makes extensiveICs in military electronics equipment, consumption for2 Is estimated at onlyillionecent reporta Sonet military-related institute, believed toigh12 there was great difficulty in procuring even theof ICsofevicesime. Estimatedand military consumption of ICs for the0sthe following tabulation:
Million VnlU
Military (reiicUiaJ) ls
PerctiU
ercent of
The output figures Idnd tbe tabulation are based conceptually on the .lumber of ICs delivered by producers (see Appendix A) and must therefore, be considered gross output. The discrepancy between output figures and identified requirements almost certainly resultsigh rejection rate hi users' production facilities. According to one source, as muchf the ICs delivered to computer manufacturers2 were rejected because they failed to meet quality standards. The percentage of rejects by military users also may be high because of the rigorous requirements of military designers for durability and reliability.
Other factors affecting the quality of output in the USSR include the lack of adequate IC test equipment and poor quality control procedures. Also, under present incentive systems, managers and workers are more concerned
ith quantity than quality of output.
w.
The Industry
Soviet IC industry is highly concentrated. Three plantslion's share of the output. Fifteen otherdrodess advanced level of technology.
IC production may exist in an additional ii plants (seend
Eksiton Plant0 workers is believed to be theproducer of hybrid ICs. The Mikron Plant, the most modem andadvanced semiconductor facility in the USSH. is the largestof monolithic devices. Much of the production line machineryin the plant is almost certainly of Western origin. The Radiois the second largest producer of monolithic ICs. and the firstknown to have serially produced monolithic devices.
(Zeleootrad)
PavtopOMd
Vinoneih
Elaiiori
Produeeii of Unhutva Slu
Fiyaano'
Kiev
Kishinev
Leningrad
L'vov
Mimk
.
MOSCOW {'ZrlenogiaJ)
Moscow
Lenin Electrical Ensfneeriai
SemVooductor
Tochelektropribor
Mno
Electrovacuum Devices Unknown
Angsoern
Komponeol
617
Serfuconducux Instruments Telephone Ecjutpoienl NuUeon
Tube end Lamp
oflana What IC Production May DnepcapeUovsl: Kherson NaJ'cbik Rybinsk Saransk Sverdlovil; Tallin
Tbilisi
Ulyanovsk
Vitebsk
YwhkarOla
known Bureau arc not
be eipenmeoUj iwodoeti* loch as ihe Vilnyus Design
of the remaining facilities have begun production ofhey are believed to be specialized producersimiteddevices for special highpriorityiauliai. and Mion in Kishinev -are relatively newdevoted exclusively to the production of ICs They probably willproducers in the future.
Manufacturing Technology
IC production processes arc backward and inefficient by
A surprisingly small portion of the production processes are automated:
Computer-aided design techniques are not used, SO far as wC know-
i
Quality control procedures and standards, and environmental control* {ventilation, cleanliness) over tourers ol contamination at crucial processing steps, are poor or non-ciistent. and
Production proceeds by rule ol-thumh formulas, with no real program evaluation anil review techniques.
The technology for quantity production of materials used InIC devices also appears to be inadequate High-quality silicon, ceramics, and special chemicalshoto realm) are in chronically short supply, even in design bureaus where requirements for tliese inputs are relatively small.
Soviet semiconductor plants operate with outdatedto some US observers, comparable to eqmpment in USoears ago.'* Although the major Soviet plants rriy on *ey Western equipment, most of the machinery in the industry is Soviet made. Indeed, much of the production line equipment appears to have been fabricated by the useractor in the wide variations in the quality of devxxs produced by different plants According to recent information, the Soviet semtconductor industry is now under pressure to obtain additional US made IC manufacturing equipment. The
" So..cl brocNuin dracriW vaiieir ot piodiMlion cQuipmrnl at modern dealgn com
pliablee art. inc|udtn| advanced Hep- andandcneiaton. multtlube diffuuo*wrluniied packaging iviteinhat jimilar to ihe US mini-mod tyMcm. Dm none ol iheic itemi hia ever been obtefved in Soviet production facilitiei. and il ii tloubdut lhat many are in wide
Sfc*>fT
Soviets suggested to one US manufacturer that, although such equipment Is embargoed, it could be shippedhird country. "sanJtiied* and re-shipped to the USSR-
A major weakness in Soviet IC manufacturing technology ii the absence of computer-controlled test equipment and diagnostic systems. Soviet testers used in both probe and final test stages are simple, manually-operated machines which are greatly inferior to Western-made system* Soviet-made testers appear to be copies of older Western systems. Somef visitors to the USSR have allegedarge proportion of Soviet test capability consists of illicitly acquired Western machines.
esult of these deficiencies in the production process, yieldsercentage of usable ICi. after full processing, are extremely low in tbeof ICs in the Soviet Union. One source
revealed that the best yield onICs achieved as of2nd dial the highest yield claimed by any IC facility producing monohlhicsat the Western-equipped Mi lion Plant iny comparison, yieldsre common in the United States in the production of the same class of monolithic devices.
Foreign Assistance
Cooperation with Eastern Europe
Aid from abroad has been instrumental in building the Soviet IC industry. Cooperation between the USSR and Eastern Euiope in semiconductor (ethnology has been mostly one-sided The OSSR has benefited from the results of RAD in Eastern Europe but has been less free in sharing its own technology Only Bulgaria and Poland arc known to have received any direct Soviet assistance.. Bulgaria received machinery and equipment for the production of discrete semiconductor devices (tTamuloes andoland was given machinery for the production ol transistors and certain types of ICs
The more industrialized countries in Eastern Europe had managed by theo advance beyond the USSR in selected areas of IC ir-chnology For example, Czechoslovakia had begun pilot-scale production ol TTL ICsear before the USSR, and nccitly Hungary may haveiglser level of cleveioprnent than the USSRrrrp)antibXM technologyS/LSI.olanda,or step toward the creation of the largest and most modem IC production capability in Eastern Europe. It contracted with SESCOSEM of Franceomplete turnkey facility (mschinery and technology) for the production of bipolarr-vices. linear ICs, and epitarial planar transistors.
h* Polish SF5COSEM contract was quickly followedovietwith Poland for joint development and production of IC devices,ew Soviet policy of closer and mutual cooperation with Eastern Europe in semiconductor technology. The new Soviet initiatives probably were encouraged by the pros|iect of gaining access lo Western IC technology.
SI. The French IC. facility was approved by COCOM infirst time lhal COCOM approved ihe sale ol Western IC technology to anybecame operational this year It will turn out nn estimated
7 million ICsn addition, through arrangements (both legal and illegal) with other Western firms. Poland has. or woo will have, the capability to produce high-powcr silicon controlled rectifiers, thick-film hybrid ICs. poly-and monocrystallinc silicon, and double-sided and multi-laycr printed circuit boards.
USSR also has cooperative IC development and productionwith Bulgaria and
t<ra -r- nowmall capability to produce
MUi 1Cs' though currerrfoutput does not appear to be sufficient to satisfy even domestic requirements. MOS ICs are of intense interest to the USSR. Bulgaria is seeking further Western assistance for MOS production and also wants, to purchase MOS/LSI devices. Hungary, and the USSR are jointly tryingevelop ion-implantation production technology, which is now emergingajor new process for large-scale production of high-density MOS and bipolar devices. Hungary is looking for Western help In this area and in the expansion of semiconductor production into more advanced areas of ICas well.
Cooperative agreements in semiconductor technology exist withand East Cermany. although the specific areas of cooperatioo are not known. The Soviets may be working with East Germany in the design and manufacture of semiconductor production machinery. East Cermany is the only country in Eastern Europe that designs andull line ofproduction machinery, much of it for export to the USSR. While East German production equipment Is markedly inferior to US equipment, improved new products are now at the prototype stage.
Levels of output of ICs in Eastern Europe are still too small (at mostillion deviceso permit large exports to the USSR. Poland now may beosition to supply the USSR with small quantities of high-quality, high.reliability devices useful for special-purpose military applications. In the next year or two, as Polish IC production reaches capacity levels provided by Ihe French plant. Poland may be able to supply the USSR with enough ICs to aid substantially in the Soviet pioduction ol third-generation computers.
The Special Case ol Yuzotlatxa
probably has given
a sizable uoost to the Soviet IC program by reexporting semiconductormachinery of US-origin to the USSR.rovision of the Soviet-Yugoslav Trade Agreement. Yugoslavia is to provide to the USSIt "special technological equipment for the manufacture of semiconductors and integratedalued atillion. Yugoslavia, however, does notsemiconductor production machinery and tacks the specialized machinery and technology needed to manufacture such equipment. Moreover, by theof senior officials of Isloa. one of the largest Yugoslav producers ofYugoslavia plans to continue to import, rather than produce,production machinery.
mports from the United States3 amounted toillion, including S3 bonders,iffusionchamberann
'! Cabled by French authority on ihe basis ollo- yield .ale. Much higlserillion units pe. .ear-ought lo be possible as Polish vieUU apn.oacli Western standards
and repeat came,m ihe United State, alone would permit YugosUvia to produce at kailenucooductors ol various type, annual^ Yugoslavia, semiconductor leqwemcna are currently estimated at aboutillion toillion unit, annually.
Weattrn Contributions to tht Sovin IC Program
oviet IC manufacturing facilil.es typically include an eclectic mil
n ElTT TPmW" United
Kingdom, West Cermany. indJipaa
The cumulative value (or quantity) of Soviet capital purchases in the West is not known, but may be auumed to run into several million dollars Hie value of test equipment and other unidentified equipment purchased from Japan alonexceeded SS
The USSR has Purchased mdmdual items of production equipment usually wthout associated tr.in.ng. ma^enance. orill III III. Il' complete, integrated.induction line has been ZJ
hdTu^
andprodueoon of wrrucooductor,
oviet effort, toC manufacturingtechnologyestern countries haveens.ve and unremittinghave sought industrial ilem, ac.cn, the rntire ipectrum of IC manu-
"ST1T?roduction lines and complete turnkey facilities. In addition, the Soviets are seeking technology for manufacturing aU types of IC, including bipolar and MOS and LSI. They arcercsl3 in US
cchnology and in agreements with] ensure the continue
transfer of future US
total potenhal Soviet IHiichase, of US equipmentestimated, many Soviet feeler, for individualat upwards ofhreec.ucsr, were1diffusion0 tons of silSneFor some types of equipment, .he Sov,etsapay exorb.tantly high pricewup( above .he market price forbonders and advanced step and repeat camera,
Prospect*
Soviet IC program is at the crossroad, II. a* in the past, themostly on its ownear term prog/eu along icveral fronts i, likely
Io bo moderate a( best, First. ou(|iut can be expanded by adding new production linn at existing IC facilities and by sating up IC production in some of tbeplants that currently do notICs. New plants also may come on stream. Existing facilities in Vilnyus and Siauhai have been scheduled to receive new plant facilitiesnd at least one now plant was to be put into operation at Zelenogradecond, improvements inyields, especially cf row-density devices (DTLs and TTLs) can be expected as managers and workers gain experience with this complex technology Yields probably will remain low by Western standards and production costs high, unless the Soviets arc able to assimilate advanced production technology more rapidly than in the past. Third, Soviet state-of-the-art in advanced seriallybipolar; LSI devices is unlikely lo reach current US levels by the. By lhat lime the lechnology gap could be enormous. MOS/LSI ICs in the United Stales currently hnvr density levelsits per chip; Ihcse arc expected to rise00 bits per chip
til If. on the other hand, the USSR could acquire IC technology and equipment from the West, progress in both the develo|>ment and production of ICs could be rapid, depending upon ihe nature and extent of WesternThe USSR could obtain Western help in three wayii
Large-scale purchases of IC product sou and test equipment for the rnoderniratson of existing plant facilities Under this option,yields and the quality of ICs now in production could beapidly though probably not lo Western standards unless related know-how also were included.
Purchase of turnkey plants and leclsnology lor the production oftypes of ICs, especially MOS/LSI.
Joint RAD programs withS find oilier foreign firms.
I mplcntentation of any of these options wouldm(or relaxation in embargo controls
n the shoil inn, Ihe Soviets me likely to view the first option nv tlvcir most uigcntige-scalr production of high quality, high reliability IC* ol SSI complexity would permit ihe USSR to design and build complex military etccfiomcs hardware to satisfy most, and ptcroabiy all. military requirements for seveial years to route, including those of the most advanced Soviet weapons systems In addition, the USSH probably would be able to initiate massol third-generalion computetl lot Civil uses.
n the longer run. la kins; up theption would permit ihe USSR to closeite current leehisnlogywith ihe United Slates Tin-third option would enable list- USSR lo stay abicui of US advances in ihr sl.ile oflhc-.it. In addition, jointiiis could help the USSR ovenotne institutional borrim rh.it currently delay gelling new products out of ihe Ijlwua-luiy and into imiusliial scale production Some iomt programs alicady .lieI lot in llu- scientific and tecluiic.iltrcenilyrw Lngu Western clcttroniia piodueets*0
SECfTET
- idiutu inc
text,ubjectigh refection rate in Ihe users* facilities.0 output figure was givenoviet manageremiconductor plantigh-ranking official of the Soviet semiconductor industry and seems reliable. The reliability of2 figure, which is based on fragmentary plant data, is less certain. Output of ICs (hybrids and monolithics) at Voronezh2 underhift production conditions is estimatedillion units. One source reported that Voronezh was one of three plants having approximately equal levels of oulpuf thatmajority- of total Soviel IC deviceshus, total output of ICs was on the order ofillion units or lew.
Quasi-official data also were provided by the same source for planned Soviet output1he Soviets had hoped to produceillion loillionndillionillion deviceshese dala am considered an unreliable guide to actual production:
Accurate planning probably was not possible in the early years of this technologically complex industry Yield* of ICs are especially unpredictable because of their extreme sensitivity to differences in worker skills, production know-how, and the quality of manufacturing equipment. Even verychange in yields are magnified into Urge changes in usefulence,eems likely that planners grossly overestimated achievable yields in projecting output1
At levels aboveillion units per year. ICs almost certainly would begin to appear In consumer and eoinmeieial end-products We have seen httle evidence of Hits.
The production of ICs depends heavily upon the uninterrupted flow of high-quality materials from suppliers outside the industry, over which Ihe industry could excit little, if any. control. In fact, the uneven delivery of materials of varying qualityajor factor impeding production
BLAMK PAGE
asic Terminology
Integrated circuits (ICs) arc miniaturized assemblies in which bothcomponents and their mteixonnections are formed by diffusion,or deposition of materials into orommon substrate. Electronicin an integrated circuit normally include both active (transistors and diodes) and passive (resistors, capacitors, and occasionally inductors) types. Most ICs may be classified as monolithic and hybrid In monolithic types, active and passive components and interconnections are formed svithin or upon asubstrate, usually silicon. Monolilhics are classified as SSI. MSI, or LSIr large-scale integration) types, in increasing order of circuit complexity and number of components per IC. In hybrid types, passive components are formed as films on an insulating substrate, such as ceramic, and separate active components are attached by bonding.
The monolithic process lends itself to high-volume production and the manufacture of complex structures. The hybrid process tends lo be limited to low-volume applications and the production of less complex structures or circuits with special requirements (such as highoutput or high operatingbeyond the reach of monolithics. Ccnerally, manufacturing techniques for hybrids are simpler than for munolithics. although in scene cases individual active componentsybrid may be high teehrsole-gy products
Hybrid ICs are of two types: thin-film and thick-film. In the thin-film type, passive components and interconnections are deposited as films by sputtering or evaporation. In the thick-film type, passive components and interconnections are formedilk scicen process Ihe thtn-film processigher degree of precision in dimensional tolerances and tends to be used svhen small packagingicrowave frequencyequired. The thick-film process is more appropriate, when small packages, tight dimensional controls, and high-frequency capability are not required.
Monolithic ICs are also of two types, bipolar and MOS. Bipolar devices are so-called because they use both negative (electrons) and positive (holes) charges as carriers of electrical current. MOS (metaloxide semiconductor) devices, by contrast, uses cither electrons (nMOS) or holesonolithic IC that incorporates both pMOS and nMOS is called CMOS (complementary MOS) Both bipolar and MOS devices require the same kinds ol manufacturingbut employ different manufacturing techniques. Dipolar types weie the first to be developed; MOS IC technology is of recent origin, sinceS Many families of bipolar ICs. each uiih variations in circuit design ortechnology, have evolved over ihe years, notably (in ordei ofresistor-transistor logiciode-transistor logicransistor transistor logicnd emitter coupled logic (ECL) Cunrntly. mostly IT! and ECL arc used for new equipment design
ECL. CMOS, and TTL are wrdoly used in general-purpose logic elements. ECL is the (attest. CMOS consumes the least power, and various types of TTLradeoff between these two Ielectrical characteristics. TTL and the nMOS and pMOS ICs are widely used in memory or other circuits in which the IC. or major segments of it. is composed of huge numbers of identical circuit elements ln general. MOS structures air less complex than bipolar equivalents, permitting denser packaging of components. Hence, when used as memories, they have higher capacity or require less space for the same capacity. Bipolar devices are faster.
Cs may be classified as digital or linear. Digital devices operatewitch; they control the flow of current and thus are widely used to perform logic, storage, and arithmetic operations in digital computers and similarLinear ICs respondontinuously varying signal; common uses include amplifiers or voltage regulators. Many ICs have both digital and linear functions included in the same circuit.
Original document.
Comment about this article, ask questions, or add new information about this topic: