Intel has cancelled plans to expand in Europe but isn’t giving up on its “great foundry” plan, says Jason Walsh
Blogs
Image: Shutterstock via Dennis
Publishing its second quarter results for 2025, Intel’s chief executive Lip-Bu Tan yesterday took the opportunity to publish a memo to all staff laying out the company’s plans for the future.
Whether the message was really intended for staff or for the press and investors, who knows, as it was published online for all and sundry to read. The long and the short of it is that AI chips are in and overseas plants are out. Additionally, staff headcounts will be reduced by some 15%, with a goal for a total global workforce of 75,000.
The most interesting thing, however, is that Intel will not ditch plans to become a manufacturer for other chip designers but, at the very same time, is seemingly rushing to undo the legacy of previous chief executive Pat Gelsinger, the man who came up with the idea of opening up contract manufacturing in the first place.
The plan now is to slow things down on this front, Tan wrote: “We will take a fundamentally different approach to building our foundry business. Over the past several years, the company invested too much, too soon – without adequate demand. In the process, our factory footprint became needlessly fragmented and underutilised. We must correct our course.
“Going forward, we will follow a systematic approach to growing our factory footprint that’s fully aligned with the needs of our customers. We will be judicious and disciplined as we allocate capital – because that’s what great foundries do.”
In addition to the job cuts, Intel has cancelled plans to open chip plants in Germany and Poland – the former location having been chosen in a hard-found competition against countries including Ireland and Italy. At the same time, Tan said the company “remain[s] deeply committed to investing in the US”.
And yet, despite the soothing rhetoric, the actual news was more cuts. “To that end, we are further slowing construction in Ohio to ensure our spending is aligned with demand – while maintaining flexibility to accelerate based on new customer wins,” Tan wrote.
This term ‘customer wins’ is doing a lot of heavy lifting: if Intel cannot secure customers for its so-called 14A manufacturing process then it could pull out of manufacturing altogether, something that is not lost on investors.
Technological sovereignty
All of this bloodletting is happening because Intel, once an industry titan, found itself significantly behind on the chip design front, allowing competitors to surge ahead and eat into its traditional dominance. And there is no doubt that Intel needed to do something spectacular: having missed the AI boom, Intel found its lunch being eaten by competitors AMD and, in particular, Nvidia.
Oddly, that spectacular thing has now turned out to be massive cuts in spending – but it wasn’t quite like this at first. Indeed, Tan’s agenda increasingly looks like a repudiation of plans laid out by previous chief executive Pat Gelsinger to challenge Taiwan Semiconductor as the world’s dominant contract manufacturer of silicon.
Whether or not Gelsinger’s plan would have worked, we will never know, but it is worth remembering that the idea of turning Intel into the foundry of choice by opening factories in the United States and EU does hold water.
Geopolitical tensions that are now more visible as the US attempts to clip China’s wings were in fact already at work for two decades as China grew from a site for cheap manufacturing into a technological powerhouse that develops businesses that survive and thrive in low margin environments. When combined with concern at China’s growing regional military clout, the desire to ensure that semiconductors are not only designed but actually manufactured in friendly regions with relatively simple logistics and supply chains is hardly surprising.
Gelsinger himself is now openly banging the drum on US technological sovereignty. Of course he is: since being removed as Intel chief executive, Gelsinger has reappeared as the head of a venture capital firm and, consequently, wants the US government to pay for the unprofitable early research that investors can’t be bothered with.
Praising the Trump administration’s plans for a sovereign wealth fund, Gelsinger took to the pages of the Wall Street Journal last week, to say while US companies were making technological breakthroughs they needed “patient, long-term capital typically unavailable from Wall Street or traditional venture funds”.
Such a fund, he said, could support “early-stage quantum companies, help national laboratories commercialise technologies, and ensure US breakthroughs remain domestically controlled.”
It could. Venture capitalists could also do that, but in any case, there is nothing wrong with the concept of a sovereign wealth fund, and the real point here is that the growing recognition that countries such as the US and blocs, such as the EU, can no longer rely on the goodwill of China or, indeed, each other. In such a context, ensuring that critical technologies – such as semiconductors, for instance – are not only designed but also manufactured within secure, friendly borders is becoming an undeniable strategic necessity. The big question now for Intel is will it be that manufacturer?