The EU will be spending €43 billion to reinvigorate its semiconductor industry. While that sounds like a lot of money, it pales behind the $430 billion invested by South Korea and the $280 billion earmarked by the US.
Europe needs to target its spending.
The final points in the negotiations were over how much money should go to research and development and manufacturing, and how much funding should come from national governments or a common EU purse. These are not the most important issues. Europe should focus on its strengths, which include imaging, advanced packaging, and power devices. Reinforcing these sectors requires both R&D and manufacturing.
What Europe should NOT do is invest billions in building giant foundries to manufacture the world’s most miniaturized chips, currently at four nanometers and shrinking. Its subsidies are insufficient. Competition is fierce. Most European industry, led by the auto industry, consumes basic and “smart power” chips with larger feature sizes so there’s little security advantage to be gained by going into the super top end. In my opinion, it’s a fantasy to pursue a dream of doubling Europe’s market share to 20% with just a €43bn investment, up from about 10% today.
Subsidies should instead reinforce significant European competitive advantages. Begin with imaging. The Netherlands’ ASML makes the world’s most advanced lithography machines. Belgium’s IMEC and a slew of German optics companies led by the famed Zeiss Group provide both cutting-edge research and a world-class supplier network.
Another key European edge is advanced packaging, the process by which multiple semiconductors are stacked onto each other. Germany’s Micro-Systems Engineering is a world leader, and French-German STM Microelectronics and Germany’s Infineon have considerable expertise. Both companies also boast cutting-edge skills in producing power devices that manage a wide range of electric current and voltage in the most energy-efficient way.
Outside the EU, Switzerland has an ecosystem of ultra-low-power expertise led by EM Microelectronic stemming from its rich heritage of watchmaking, while the UK has enormous and world-leading design expertise at Europe’s largest technology start-up cluster in Cambridge.
A key variable is Germany’s attempt to woo the US semiconductor giant Intel. A delegation of the Saxony-Anhalt government, the German region where Intel plans its €17 billion manufacturing plant, flew last week to the US. Intel reportedly wants additional subsidies before committing to the project.
This investment may not be sustainable. It seems unlikely that such a megaproject ever could be competitive in a high-cost country such as Germany without ongoing subsidies. If Intel invests in advanced manufacturing in Europe, it would be much better off in Eastern Europe, such as Poland or the Czech Republic. Both have lower costs — and yet a highly educated and skilled labor force.
This is a key debate: will Germany and France get the lion’s share of the subsidies just because they are the richest and biggest EU economies? Since most of the money will come from national budgets, it looks like Germany could become the big spender and the big beneficiary — to the detriment of the more competitive, but smaller and poorer countries? These are political questions outside my semiconductor expertise. But the risk remains that Europe could invest in the wrong places.
Another wild card is Taiwan. The country’s chip manufacturer TSMC is expected to soon give more info on its European plans. It could invest in Eastern Europe — or decide that it is not worth investing in Europe, with or without subsidies. The company already has a 20% world market share and 90% for the most advanced four and three-nanometer chips. The fabless chip giants Nvidia, Broadcom, and Qualcomm depend on TSMC, and the company is building a new multibillion-dollar fab in Arizona.
From a business perspective, TSMC doesn’t need to make a similar investment in Europe. It will only go ahead if it receives maximum subsidies. And a TSMC plant in Europe will probably not make the most advanced chips — these will stay in Taiwan so it can maintain its Silicon Shield and keep leverage against a Chinese invasion.
My own country, the UK, risks losing out, another casualty of Brexit. It will not participate in the new EU chips program. We have been waiting a considerable time for the UK government to announce its own semiconductor strategy. Semiconductors require scale and the UK simply is too small to count by itself.
What a pity. The UK packs a considerable punch on semiconductors. ARM is one of the world’s most advanced chip designers and the undisputed leader in processor licensing. It was born here in Cambridge, where I live and where there is now a significant semiconductor ecosystem. Another impressive chips cluster is in Wales. It has pioneered compound chips, semiconductors based on other materials than silicon, such as gallium nitride. And the UK is a world-beater in sensors.
The best way forward would be to support these UK standard bearers – and help them to begin to manufacture in Europe. The lower-cost high-skill economies of Eastern Europe could reduce outsourcing to China. Unfortunately, the present Brexit-strangled UK government would not find it politically palatable to subsidize operations on the continent or even participate in EU research initiatives, judging by the continuing standoff over participation in the Horizon Europe program.
A final word of advice — make sure the EU’s chip subsidies go to innovative SMEs and start-ups, not just the large companies. We must help vibrant newcomers grow, not just pour money into big old giants and fall back on retrograde thinking, trying to turn the clock back to a past seen through rose-tinted spectacles.
Remember what happened to the British car industry? In 1951, the UK was the world’s second-largest manufacturer of autos, with famous brands such as Austin and Triumph. Management mistakes drove these world-famous companies to seek state support. The subsidies allowed them to become inefficient. British car companies lost market share. They consolidated into a single state-owned company — until they were shut. Today, we have no domestic car industry.
Europe must do better with chips.
Christopher Cytera is a non-resident senior fellow at CEPA. He is a technology business executive with more than 30 years of experience in semiconductors, a veteran of successful start-ups such as Alphamosaic and Spectral Edge as well as industry giants such as Broadcom and ST Microelectronics.
Bandwidth editor and CEPA non-resident senior fellow Bill Echikson helped research and write.
Bandwidth is CEPA’s online journal dedicated to advancing transatlantic cooperation on tech policy. All opinions are those of the author and do not necessarily represent the position or views of the institutions they represent or the Center for European Policy Analysis.