Intel has officially entered the grin and bear it phase of its recovery

Comment On Monday, Intel's share price surged on word it was spinning out its foundry biz as an independent subsidiary and signing AWS and the DoD as customers. But while Wall Street celebrated, Chipzilla's road to recovery is far from over and everyone involved, whether they are shareholders, employees, or partners, are going to have to grin and bear it, either until CEO Pat Gelsinger can realize his vision or its board cuts his tenure short. Some are already feeling this more than others.

Along with the spinoff, Gelsinger announced sweeping changes to the organization's structure, consolidating its networking and automotive groups with its client division, and, perhaps more concerningly, pausing development of its German fab and Polish assembly sites for two years. The announcement has cast doubt on the future of the facilities, not to mention the EU's goal to double its share of semiconductor development, manufacturing, and material supply chains from 10 to 20 percent by 2030. EU member nations have already put up €43 billion ($48 billion) in subsidies to support this goal.



Intel was originally slated to receive roughly €12 billion ($13.4 billion) in state aid between its German and Polish developments. It seems unlikely that aid will ever materialize and, without it, Intel's two-year delay may end up being permanent.

The timing of the announcement was particularly awkward because it came just days after the Polish government received the green light from the European Commission to fund Intel's planned test and assembly facility. Yet even with the funding, these facilities would have been exceedingly expensive, with Intel needing to come up with more than €20 billion ($22.3 billion) in capital, at a time when it is aggressively cutting costs and headcount.

By the end of 2024, Intel will employ nearly 16,000 fewer workers around the globe . Intel was one of just a handful of leading-edge foundry operators, and unless someone else steps up to fill the void, Europe could very well be relegated to producing little more than museum-grade chips . Samsung is certainly large enough to step in, but the economics of doing business in Europe may not be attractive enough even with a few billion in subsidies to take the edge off.

While Intel is very clearly prioritizing its US investments, the spinoff for all its fanfare won't stanch Foundry's bleeding, even with AWS and the DoD lining up for their share of capacity. Announced alongside the spinoff, AWS confirmed plans to manufacture an AI fabric chip and said it would commission a custom version of the x86 giant's upcoming Xeon 6 processors. Meanwhile, Uncle Sam has committed $3 billion to Intel to establish a secure supply chain of semiconductors under a program known as the "Secure Enclave.

" While a much-needed win for Intel Foundry, it'll be some time before Intel can realize any revenue from either deal. The fact remains that Intel's foundry business is struggling to bring in a profit – the group reported $2.8 billion in operating losses in the most recent quarter – and this is unlikely to change anytime soon as an independent subsidiary.

So, what exactly is getting spun off? By all accounts, not all that much. Under the new structure, Intel Foundry will operate as an independent subsidiary inside Intel. This means it'll have its own board, more autonomy to pursue new sources of funding, and potentially better optics among fabless chip companies wary about building chips at a competing firm.

However, even compared to Intel's spinoff of Altera late last year, Gelsinger is keeping a very tight leash over the division. Instead of electing a new chief executive to oversee the company, Intel Foundry's staff will continue to report to him. This will no doubt change with time, and we expect an IPO is somewhere on the horizon, but for the moment, Intel's product future remains deeply entwined with Foundry's success.

With the decision to move its Arrow Lake CPUs to TSMC, very little of Intel's 2024 product portfolio is manufactured in-house anymore. However, if everything goes to plan, that'll change as production of Intel's 18A process node ramps in 2025 and reaches volume production in 2026. Intel really needs this plan to play out without delay.

Unlike TSMC, which has hero customers like Apple, Nvidia, and AMD to support it, Intel remains Foundry's biggest and most important customer. When Gelsinger first returned to Intel as CEO in early 2021, he wasted little time setting forth an ambitious plan to reinvent the ailing chipmaker. This go-big-or-go-home strategy was solidified when just over a month after taking the helm, he surprised many by announcing his intention to open Intel's fabs to contract manufacturing and invest $20 billion – a sum that's since grown to roughly $30 billion – in two new leading-edge fabs in Arizona.

In the years since, Intel announced the development of new "angstrom-era" process tech, committed two additional fabs in Ohio at the cost of roughly $20 billion, enlisted the help of private equity firms like Brookfield and Apollo to finance the projects, and secured $11.5 billion in government subsidies under the CHIPS and Science Act. In doing so, Gelsinger has made it incredibly difficult to abandon its foundry division.

Brookfield and Apollo have some $26 billion wrapped up in Intel Foundry's success. Meanwhile, Uncle Sam is even more dependent on Intel. After GlobalFoundries abandoned the development of its 7nm process node in 2018, Intel became the US's only domestic supplier of leading-edge process technology.

This arguably makes the company the single most important chipmaker in the country and deeply intertwined with American national security policy. Put bluntly, the US can't afford for Intel Foundry to fail, especially considering the gains China has made in spite of ever tighter trade restrictions. And even if Intel did try to sell off Foundry, who would possibly be crazy enough to buy a company that’s losing billions each quarter? For Gelsinger, this gamble is Intel's biggest transformation since it transitioned from memory to microprocessors, and that may be true, but nobody said anything about this being easy.

.. or fun.

®.