Intel can romance European authorities to fund its fabs until it sees blue, but what it needs most is to dispel the suspicion that its advances are more about money than about love.
Car manufacturers are worried about the chip shortage now, but they should also worry about their chip supply ten years from now, Intel CEO Pat Gelsinger warned during a keynote speech at the IAA auto show in Munich. He predicted that semiconductors would make up 20 percent of a premium car’s bill of materials by 2030, up from 4 percent today. “Semis become central to every aspect of innovation and the supply chain for cars. As important as it is today, you ain’t seen nothing yet.”
Conversely, said Gelsinger, automotive is becoming increasingly important to the semiconductor industry. With a projected total addressable market of 115 billion dollars in 2030, automotive is leaving behind its niche status in the semiconductor realm. And that’s why Gelsinger traveled to Munich: to woo the European car industry. “We need you and you need us.”
It makes sense for Intel to make automotive a key strategic priority. But cozying up to Europe’s industrial crown jewel clearly served an additional purpose: sending a message to European and national authorities, spurring them on to subsidize Intel’s fabs. The processor maker seems to have dropped all the subtleties in its courtship lately, to the point that it’s looking a little bit too eager.
In his presentation, Gelsinger sketched out what Intel’s IDM 2.0 strategy would look like for the European car industry. He argued that automotive chip designs will increasingly depend on leading-edge silicon, and made clear that Intel is eager to manufacture these devices, in close proximity to those who buy them.
To that end, Intel announced that it’s committing 16nm foundry capacity at its fab in Ireland and launching services to help automotive chip designers to transition to advanced nodes. Under the Intel Foundry Services Accelerator program, Intel will open up its IP to customers and set up design teams to support their efforts to draw up the blueprints for the next generations of automotive-grade chips. Eventually, Gelsinger said, these devices will move to ‘Intel 3’ and ‘Intel 18A’ nodes, which will commence production in Q3 2023 and Q3 2025, respectively.
Reportedly, European companies are enthusiastic about Intel’s initiative. According to Reuters, nearly 100 automakers and key suppliers – including BMW, Volkswagen, Daimler and Bosch – have expressed their support. It’s not clear, however, whether any of them have committed to becoming customers.
Intel’s top executive didn’t have anything new to say about its European manufacturing ambitions, though he did use more colorful language. “We wanna build a big, honkin’ fab in Europe,” Gelsinger said in an interview during his visit to the trade show. That still means starting off with two fabs, gradually expanding to eight. At 10 billion euro a piece, that would amount to an investment of 80 billion euros over the next decade or so.
Of course, construction of this megasite still hinges on public support. It’s no secret that the European Commission, several large EU member states and even local governments are willing to co-fund semiconductor manufacturing operations, but the question is whether they consider Intel – which has forfeited industry leadership – to be a suitable partner. As an established foundry and technology leader, TSMC would be a safer choice. The Taiwanese, however, are either playing extremely hard to get or really don’t see the point of building a leading-edge fab in Europe (probably the latter).
Half a year has passed since Intel made its intentions public, but the EC still hasn’t dropped the slightest hint it’s interested. Undeterred, Gelsinger sends bigger bouquets and ever sweeter messages.
Alongside to his keynote presentation, Gelsinger published an article in Politico exalting the many benefits of building Intel fabs in Europe. They’ll supercharge the continent’s high-tech ecosystem far beyond semiconductors, the CEO wrote. “Wherever semiconductor technology clusters have grown, so has the pool of innovators, entrepreneurs and visionaries that have then gone on to build new companies powering brighter, greener digital futures.” If all the pieces come together, “this will be a magical European success story,” he told Fortune.
Gelsinger could personally serenade European Commissioner Thierry Breton, there will always be a suspicion that Intel’s advances are more about money than about love. Decision makers are well aware that Intel is in a tough spot and needs all the help it can get to stay relevant. This is not to say a marriage between the EU and Intel could never work out to the benefit of both, it’s just that there’s more risk involved than politicians are typically willing to take. Especially since the EU would be subsidizing a ‘foreign’ company.
If Gelsinger wants to move things forward, he needs to make a grand gesture. It’s unrealistic to suppose that the Eurofab would only supply chips to the European market – in fact, it would be beneficial for Europe to become a stakeholder in worldwide leading-edge chip production. But neither is it realistic that the EU would massively support a US company without any guarantees that European interests will be adequately served. That’s why Gelsinger needs to relinquish some control over his big, honkin’ fab.
It will be tough for Intel to, for example, grand priority access to European customers. But with the Astrazeneca vaccine fiasco fresh in mind, the EU likely won’t settle for less. And nor should it.
Main photo credit: Intel/Walden Kirsch