Why Europe Lost Semiconductors
today Europe is both a minor producer and consumer of chips EU firms make up a small percentage of semiconductor design fabrication and assembly there are a few World leading equipment and chemicals companies insert necessary mention of asml and St microelectronics and Infinity on here but that's about it none of this was inevitable Europeans independently invented the first transistor just a few months after the Americans Europe France in particular was an early research pioneer but this early optimistic lead was squandered by 1960 even the Japanese produced more transistors than the French the European semiconductor lag has persisted in the decades since can the EU build a semiconductor Airbus is the EU chips act the right way forward we can look at those questions later for now let us start with the very first question how did Europe fall behind in semiconductors in the first place 50 years of History right here and note per what happened in 2016 I'm not including the United Kingdom in here I guess they'll get a separate video later thanks brexit first let me talk about the asianometry patreon Early Access members get to see new videos and selected references for them before their release of the public it helps support the videos and I appreciate every pledge thanks and all with the show in 1947 Bell Labs created the first transistor but soon after announcing their invention six months later in 1948 word came of two scientists in Paris inventing something similar it was all true during World War II Herbert Matera and Heinrich Welker studied semiconductor materials to make radar receivers for the German radar program the Germans were attempting to catch up to Allied radar systems Matera worked at telefunken's Berlin Labs as the Soviets approached near the war's end the lab relocated its equipment abandoned to Central Germany wilker worked at Munich University throughout the war studying the electrical properties of high Purity germanium his work interrupted when Allied bombing destroyed his lab in October 1944. for those unaware Germany lost the war after that Matera and welka relocated to France as they couldn't do research in occupy Germany they resumed their work in semiconductor materials in Paris at a subsidiary of Westinghouse their first task was to set up a production line for mass-producing germanium rectifiers for the Telecommunications and Military Industries rectifiers are electrical devices used to convert AC power to DC power on the side they studied a curious electrical effect that Welker had noticed and experimented with before the war ended they built a Contraption with two electrodes held very close apart about 100 micrometers the two electrodes touched a sliver of high Purity germanium they then found that if you vary the voltage in one electrode you can control and even amplify electrical fields in the other electrode despite the two electrodes being physically apart this contraption is now known as the point contact transistor and it is conceptually quite similar to what bell Labs originally developed in 1947. has the two discussed their Contraptions Behavior Bell Labs publicly announced its invention of the transistor in 1948 and with that The two scientists realized what they had they showed it to France as Secretary of post telephones and telegraphs Eugene Thomas who urged them to patent it Thomas also suggested the name transistron in May 1949 Thomas held a big press conference announcing the transistrons invention he cited the transistron superiority to vacuum tubes and demonstrated its use for higher quality radio receivers the achievement was lauded in The French Press calling the transistron superior to its American counterpart patents were granted for the transistron in both the United States and France in 1951 and 1952 respectively in late 1949 Westinghouse built a factory in France capable of producing tens of thousands of transistrons each month for use as amplifiers in the telephone system but shortly after the press conference the French government with thoughts of Hiroshima burning in its mind decided to focus its resources on developing nuclear technology the Telecommunications industry was left a Private Industry without Government funding work on the transistron languished and its inventors eventually had no choice but to move on welka returned to Germany where he worked in Siemens r d Labs there he produced some pioneering work that paved the way for today's LEDs he retired in 1977 and passed away in 1981. Herbert Mater also returned to Germany where he founded a small semiconductor production company intermetal which he eventually sold after that he immigrated to United States Malibu California actually great place to retire he studies semiconductor materials into his 90s before passing away in September 2011 a few weeks short of his 100th birthday he lived long enough to be publicly recognized for his role in semiconductor history a life fully lived indeed Europe's semiconductor efforts in the 1950s centered on France West Germany set out much of the early part of this decade due to technology restrictions imposed by the Allied occupation Francis early semiconductor achievements by 1949 were impressive they not only included the aforementioned transistron but also a replication of the original transistor however the French effort constantly struggled with manufacturing the early years of semiconductor manufacturing were wild messy and interdisciplinary teams struggle with the unpredictability of early semiconductor manufacturing and it caused them to conservatively turn towards more abstract and theoretical research the American Semiconductor industry corralled this early unpredictability there were a number of reasons how first early transistor production knowledge quickly dispersed out of bell Labs into a variety of other companies Bell Labs held open Symposium to detailing the production knowledge and liberally granted licenses they did this partly because of Anti-Trust concerns but they also believed in spreading around the knowledge one VP said if this thing was as big as we thought we couldn't keep it all to ourselves it was our interest to spread it around if you cast your bread on the water sometimes it comes back angel food cake what on Earth is angel food cake anyway attendees and licensees included vacuum tube incumbents and electronics firms many licensees contributed new Innovations of their own including Texas Instruments silicon Junction transistor and diffuse transistor second American teens were more interdisciplinary than French teams take the Traders eight the co-founders of Fairchild semiconductor originally recruited by William Shockley Shockley had been one of the transistors original inventors back at Bell labs chocolate deliberately recruited an interdisciplinary group of them only Robert Noyes had a formal semiconductor background the rash trained in physics mechanical engineering or Optics and third Shockley also pushed his scientists to handle technological processes themselves so that they can see and devise new ways to improve production take Dr Jean ornee a Swiss American who was one of the eight his experience perfectly illustrates these principles at work he trained in mathematics and physics not electrical engineering and while working for Shockley or knee had the chance to hang around the semiconductor shop floor watching and building up familiarity with a rapidly evolving processes of semiconductor Manufacturing he frequently sketched out ideas on notebooks on December 1st 1957 he sketched out the ideas of what would eventually become the planar process he later took this with them to Fairchild I won't get into the technical specifics though I want to what you should know is that the planar process allowed Fairchild to produce reliable transistors at high scale another reason for France's turn towards the abstract was a lack of joint business government initiatives bringing theoretical breakthroughs to the market take the French semiconductor research initiatives unofficial Patron the French physicist Pierre igrane igrane trained in the United States and returned to France to found the solid state physics Lab at ikol normal Superior a research University in Paris igrane had a wide range of interests but industry was not one of them he taught a generation of semiconductor researchers and most of them remained in the theoretical space on the other side French businesses under invested in r d a survey of 34 French companies from 1880 to 1940 found that under a quarter had in-house r d Labs the French electronics Industry continued to use outdated but familiar vacuum tube technology perhaps unaware of what transistors would eventually do to the trusty tube later on transistorized French radios did not widely appear until around 1958 very late in the 1957-1958 period France experienced a series of Economic and political crises partly due to the Algerian War this instability caused the then Finance Minister Antoine Piney to implement a wide-ranging austerity policy the French transistor industry was at the time 90 dependent on the government including pending initiatives to develop produce and sell transistorized television sets without that government investment in french-made electronics components friends consumer electronics companies turn to Imports but with the Frank currency falling in value at the time again due to various political crises margins shrink this eroded profits and impeded future investment a punishing cycle so by the end of the 1950s the French semiconductor industry had squandered its early research science lead falling behind Japan and even its peers on the continent the 1960s saw the rise of the integrated circuit or IC the IC was joint invented back in 1958 by Texas Instruments and Fairchild semiconductor so they controlled it at the start where before Engineers carefully made and optimized a discrete circuit design as it was being made the IC was a black box in which all the decisions were already made here we should acknowledge the military's critical role in crafting the direction of the American Semiconductor industry by the 1960s the military bought half of the semiconductor industry's total output projects like the Minuteman 2 ICBM which began in 1962 drove the need for more powerful guidance computers they justly turned to ICS just a few years after its invention that's a big deal not to mention the Apollo program between 1961 and 1965 the Apollo program was the single largest consumer of integrated circuits though the exact number they bought varies depending on your source this early American Military demand exerted a strong influence on the industry's direction the military cared for devices that were smaller useless power and above all were reliable in other words they wanted higher performing devices rather than cheaper devices forcing American Semiconductor firms to eventually develop and hone the techniques for such devices by contrast in Europe semiconductor demand in the 1960s came from the consumer sector rather than the government slash military for instance the military made up 50 of the U.S semiconductor Market in 1960 but it was just 29 in 1963 France and just eight percent in both 1963 West Germany in 1963 Italy by contrast in 1963 consumer made up 56 percent of semiconductor demand in West Germany 35 percent of France and 50 of Italy it was just five percent in the United States in 1960. consumer demand is different from military demand in that it places a higher emphasis on price European electronics companies like Phillips and Siemens were vertically integrated Giants who viewed semiconductors as a cog in their final machine rather than an output in of itself as a result these firms focused on evolutionary improvements to their in-house transistor Technologies for instance Phillips's post-alloid diffused transistor an improvement to the germanium alloy Junction transistor these transistors have to be individually manufactured and suffer mechanical limits on how well they can perform but they were cheap which was What mattered at the time Phillips and Siemens believed that silicon would only ever be suitable for high temperature applications like power semiconductors so they only focused on improving their germanium transistors over in the United States the industry was peppered with startups and spin-offs companies like Shockley semiconductor Fairchild Intel and AMD produce semiconductors as an end product these companies discovered silicon Valley's more disruptive Technologies once Fairchild started mass-producing silicon planar transistors it obsoleted the entire alloy Junction transistor space widening the gap we can't blame Europe or any Nation for not coming up with something revolutionary lightning can strike anywhere what more matters is if you can catch up Japan started in a similar position to France or Germany but the Japanese consumer electronics Industry rapidly adopted transistor technology and caught up products like Sony's transistor radio at first grew on domestic demand the Japanese government encouraged radio and television adoption throughout the 1950s and 1960s by the mid-1960s 85 percent of households had a TV and there were 1239 channels the government also helped by gating Imports to Japan allowing its industry an Eclectic mix of big and small firms to build semiconductor manufacturing competence Mass semiconductor production kicked off in 1955 and grew 125 times in the second half of the 1950s matching American Levels by 1962. items like
transistorized radios and the color TV set became massive hits overseas to advance our semiconductor proficiency Japanese company signed technology licensing agreements with foreign companies there were over 250 such agreements from 1952 to 1968. they quickly mastered the Acquisitions and by the mid-1960s they were producing their own Innovations on top of them a mark of successful transfer and while France and the other European nations got companies like IBM and Texas Instruments to build factories on European soil in the 1960s successful technology transfer never happened perhaps because of more limited labor Mobility Trends in Europe or because they never asked for it in 1961 the French government set up a plan to support its domestic calculator and computer Industries this effort turned France's computer industry against its own nascent semiconductor industry when the latter attempted to Garner State support for themselves they were rebuked by the computer guys who said that they could not find French semiconductors competitive in quantity nor quality ouch then in 1966 France launched its computer development program planned calcul this was in response to various external factors IBM's increasing computer dominance in Europe the American government's 1963 refusal to allow computer sales to the French atomic energy Commission and General Electric's 1964 acquisition of France's biggest computer maker plan calculus sought a national computer as quickly as possible with this has a primary goal a domestic French semiconductor industry slipped to the back burner where languished I should emphasize that the French semiconductor Gap had little to do with any mind-blowing Innovation the mid-1960s were still early the IC had been invented less than a decade ago it was really all about production again French chip makers could not gain sufficient production scale to keep up with their foreign competitors aggressive cost cutting in 1964 fairchild's Bob Noyes decided to sell the IC at a price lower than that of an equivalent discrete Circuit of diodes transistors and resistors the price was lower than the manufacturing cost but noise was betting that the price cut can stimulate market demand and with that they can figure out some way to make the books balance later noise turned out to be right A year later in 1965 Fairchild received an order for 500 000 ICS a year after that in 1966 the computer maker Burroughs bought 20 million Fairchild ICS scale and Moore's Law figured out the rest European semiconductor manufacturers survived in Niche parts of the market like analog circuits again Phillips in the late 1960s they developed an analog signal receiver IC for their television products Phillips along with Siemens remained strong in this analog niche the transistor is an integrated circuits of the 1960s were simple enough for industrial catch-up to be feasible this change in the 1970s in 1971 Intel ushered in a new era of rapidly increasing complexity with the introduction of the first microprocessor the Intel 4004 designed by the italian-born and naturalized U.S citizen Federico fagin previously logic and math operations were handled by discrete chips created for that specific purpose which was kind of like Reinventing an engine every time you wanted to build a car the special thing about the microprocessor was how it Consolidated together those functions into a single swappable package it takes in data inputs from memory executes instructions using those inputs and then sends the outputs to memory very rapidly microprocessors grew the number of instructions they can handle by stuffing tens of thousands of transistors onto a single chip complexity compounded on itself especially as Intel began pushing the envelope on Moore's Law this new era of large-scale systems or LSI meant that any new entrant would have to invest significant resources to make a splash the semiconductor industry was maturing in 1969 computers overtook government and Military demand as the leading demand driver consuming 44 percent of the American IC Market ERS demand semiconductors different than those for missiles electronic watches TVs and radios high performance is the name of the game pushing firms to create more technically aggressive products Europe's big flaws were not only that it lacked the production chops to produce these high performance LSI devices they also lacked a strong source of demand for them Europe's computer companies were far smaller than that of the Americans I will discuss the legendary impotency of the European computer makers against IBM the much discussed computer Gap in a future video the other traditional semiconductor demand Industries military telecommunications and consumer electronics could not pick up the slack in the same way European military and avionics demand was pretty weak West Germany's in particular it also tended to favor a small circle of large established firms which perpetuates incumbent advantage and disregards disruptive technology and as for the Telecom market demand was quite fragmented with each of the big telcos doing their own thing so again this just leaves the European consumer electronics companies Philips Siemens and so on to absorb semiconductor demand and like I said before they want a different type of Chip that is not the Leading Edge this time Europe was not alone after a strong start in the 50s and 60s the Japanese semiconductor industry also fell behind the United States and LSI systems in 1974 Japanese IC output was worth just a quarter of the American industry 560 million dollars the 2.1 billion this situation persisted throughout the decade in 1978 Japan had 20 worldwide share of the IC Market lagging far behind the American 74 share Europe had just six percent on the critical Battlefield of the computer industry Japan needed help the Japanese computer industry was putting up a valiant fight against what the Japanese called the monster IBM but they were losing alarmed the Japanese political leadership ultimately moved to reorganize their domestic computer industry and launch the very large-scale integration or vlsi project in 1976. the vlsi project budgeted hundreds of millions of dollars to unite the country's computer makers and get them to cooperatively develop Next Generation vlsi technology the goal was to produce a one-chip computer before IBM can I covered this effort in a few previous videos feel free to watch it the project did not achieve its one chip computer goal but it did help the Japanese semiconductor makers hone the techniques necessary to put tens of thousands of transistors onto a single chip rocketing them into the generational lead combined with ample Capital via connected Banks and the industry's fortuitous technological choice to go with CMOS over nmos the vlsi project revitalized Japan's semiconductor manufacturing and Equipment Industries to compare let's take a look at West Germany West German industrial policy differed from his Continental peers in being more strongly Market oriented and rather decentralized with a few exceptions the West German government did not intervene into the market their technology sector harbored large Flagship companies like Siemens a leader in power engineering and Telecommunications equipment and AEG one of the largest electrical companies in the world in 1972 the two were West Germany's two largest indigenous semiconductor producers with 26 and 12 of the market but they lagged the Americans by a significant margin Siemens themselves estimated in the late 1960s that they trailed the Americans by a decade a gap that only widened with LSI and vlsi Technology the West German government did see microelectronics as a point of emphasis they budgeted 1.4 billion dollars from 1974 to 1979 specifically to advance their domestic semiconductor industry however this money was funneled through the gateways of their big Electronics firms from 1974 to 1981 Siemens received 1.3 billion dollars and AEG received 0.4
billion dollars this over-reliance on a small circle of corporate Giants can be frustrating Siemens is particularly conservative preferring to invest in bonds rather than technology and earning the jeer a bank with an attached electronics department then came the decline of the country's second Electronics giant AEG in the early 1970s aeg's consumer electronics business lost its competitiveness taking down the firm AEG turned losses throughout the back half of the 1970s including an Abrupt 664 million Deutsche Mark loss in 1974 and a staggering 968 million Deutsche Mark loss in 1979. amid squabbling between trade unions Banks and the government AEG was eventually sold to the French company Thompson Brandt and taken apart and all through the decade the two lost share in their own semiconductor Market Europe's largest by 1978 Siemens market share had fallen from 26 to 21 and aegs from 12 to 9 percent the European domestic semiconductor landscape of the 1980s can be best described as mixed on one hand High performing semiconductors were being made on European soil and Europe has a few large semiconductor makers but you would be hard-pressed to find a European semiconductor producer capable of challenging The Americans a decade of encouraging foreign chip makers in the 1970s to jump tariff walls and build factories on European soil only put more pressure on domestic companies like Siemens and Phillips policymakers had started to worry about their I.T industry in the 1980s Europe had opened up a large and growing Information Technology trade deficit 4.1
billion dollars in 1980 and almost 7 billion dollars in 1985. in 1985 Siemens was the only top 10 firm in Gate array manufacturing there was no European firm in the top 10 of silicon CMOS technology this deficit exists partly due to European tech companies tendency to strike technology alliances with the Americans an example is Sweden's Ericsson Erickson owns a component subsidiary called radio industry's fabric cuts the bowl lag anyways rifa producing analog radios and Telecommunications a traditional European stronghold in the 1970s rifa struck an alliance with national Semiconductor in the 1980s they moved to gec Plessy AMD and finally Texas Instruments these eventually helped Ericsson build up one of Europe's few Fabs up in Kista just north of Stockholm other examples include Philips bindsignetics and working with AMD RCA and Texas Instruments for their part Siemens struck alliances with Intel Toshiba and Fujitsu see a pattern just 25 of European I.T companies struck such alliances with another European company after spending Untold sums of money to build and support national champions in the 1960s and 70s Europe's policy makers changed directions over in France in 1982 president Francois midorandt unveiled a five-year microelectronics plan to spend some 4.5 billion francs a year to do chip r d build industrial University connections and subsidize French companies losses the goal was to win back domestic market share and small computers and semiconductor manufacturing tools quote a Big Bill but Independence comes at a price they said what they didn't know was how big the bill would have to be French subsidies drove new levels of investment but paled compared to the economic might of the Americans and Japanese in 1984 semiconductor Investments made by French companies was just 1.4 percent of the world total French companies not only struggle with labor and Technology problems they were constantly running short on funds over in West Germany the government in 1984 launched a big semiconductor r d program it would give a billion Deutsche marks over the next four years the support a variety of projects prominent was the Mega project a high-profile joint effort where the West German and Dutch governments would help Siemens and Phillips develop a four megabit dram and a one megabit SRAM respectively by 1988. they called it rather dramatically in my opinion Europe's last chance against the Japanese chip Invasion Toshiba and the rest of the Japanese were at least three years ahead Siemens completed their project and debuted their four megabit chip in March 1987 with much fanfare it was a huge crash effort between European scientists and industrialists which eventually delivered that's great but there were some caveats first behind the scenes Siemens had to license their one megabit dram technology from Toshiba Phillips Hedge 2 working with matushita for their one megabit SRAM secondly the space was moving too fast Japan's NTT had already debuted a 16 megabit dram chip by March 1987. the
Koreans were coming too with Samsung rolling up in 1989 the mega project evolved into a 4 billion chip research project called Joint European sub Micron silicon initiative or Jesse it was a pan-european Consortium to develop six megabit drams and advanced custom chips Jesse ran for seven more years but suffered serious issues Phillips pulled out from memory entirely in 1990 due to increasing financial difficulties Britain's ICL was ejected after being majority acquired by Fujitsu that same year and then Siemens shocked the European semiconductor Community by striking an alliance with IBM abandoning their plans for a 16 megabit memory device Jesse's budget increasingly shrank over the years impairing the organization's ability to function despite a long-just-stating and publicized collaboration with the US's sematech in 1991. Italy Belgium and Sweden had their own projects too notably Belgium's microelectronics r d subsidies led to the founding of iMac the world's leading semiconductor research facility the 1980s all policy goal-setting responsibilities start to move from these individual nation states to the European Community I'll cover these pan-european efforts in a future video in 1979 European producers had 60 percent of world semiconductor production that went down to 12 in 1985 and now in 1991 just 10 that year European companies had less than 30 percent of their own Market the 1990s saw technological and financial barriers to semiconductor entry get Sky High and getting higher a 250 nanometer memory Fab back then already cost about one to two billion dollars All In competition also intensified with the likes of tsmc Charter semiconductor and Samsung all of whom were located in Asia near the growing Asian Electronic Supply Chain tsmc's pioneering independent Foundry model in particular threw opened the gates ON Semiconductor Outsourcing the more fabulous companies who went with tsmc the more scale they accumulated the South Korean firms had their trebor which bankrolled their semiconductor arms initial losses so if we go back to semiconductor demand structures times continued to disfavor Europe European firms lost the consumer electronic space to Asian ones military and government spending declined after the fall of the Soviet Union and Computing remained with the American firms Europe's positions in the remaining Fields industrial telecommunications and automobile are still strong and stable but they aren't capturing the majority of the market so why did Europe lose semiconductors there was never a single reason what we have today is a situation that is compounded over half a century the US and Europe discovered transistor technology at about the same time at the end of the 1940s but in the 1950s Europe fell behind the Americans for the first time due to political technological and production issues in the 1960s strong American military spending pushed forward the industry's technological progress building the necessary skill and competence to win the market furthermore that American military spending went to smaller disruptive spin-offs and startups rather than traditional vertically integrated Giants in the 1970s has that semiconductor competence began compounding onto itself European policy focused on building bigger computers to match up with the Americans IBM incorrectly disregarding semiconductors has a future Focus Japan did not make that mistake then in the 1980s and 90s despite some very real success in r d the semiconductor manufacturing Outsourcing Trend plus the rise of the Asian supply chain intensified Market competition relegating European manufacturers to Specialty markets now we are in the 2020s in 2021 the EU announced plans to build strategic autonomy by increasing the area's share in high-end chip fabrication the goal is to double Europe's share in semiconductor production from its current 10 to 20 that market share has been around 10 since 1990 30 years history has shown that Europe's government and the EU were more than willing to financially support and promote their semiconductor Industries they remain so today but previous rounds of government money r d coordination and forced company consolidations have yet to reverse the situation well the EU chips act change course or just be more of the same all right everyone that's it for tonight thanks for watching subscribe to the Channel Sign up for the newsletter and I'll see you guys next time
2023-04-25 12:25
