Separating Intel's fabs from the design business makes no sense

Intel has had a tough time of it and has recently fallen behind some of its competitors, who would love nothing more than to see Intel fail. First of all, the company's huge capital expenditures to expand its manufacturing and supply chain — particularly CEO Pat Gelsinger's ambitious “five nodes in four years” plan — have not yet been offset by revenue growth. Worse, its guidance for the coming quarters is not as robust as one would like either. This is, of course, from an investor's perspective.

In addition, Intel faces greater competitive pressure from more rivals — starting with Nvidia, AMD and Qualcomm — than at any time since the company moved from memory chips to microprocessors in the 1980s. The fact is that under previous management, Intel shot itself in the foot years ago with some of its transistor technologies and process nodes, and it's hard to come back from that. When you have these problems, it affects your designs. For example, if architects originally planned 16-core chips, they had to move to 12 cores to meet performance limits or wafer size.

Chances are that a complete turnaround will take at least a few years. Things were already bad enough to require major adjustments, most notably with Intel's announcement a month ago that it would cut 15,000 jobs as part of the $10 billion in cost savings promised for 2025. Not all job cuts are bad, but they shake things up unless you're selling off a business unit entirely or shutting down a scientific project with no revenue or profit.

And then Bloomberg topped it off by reporting that Intel is considering a number of other options, most notably the possibility of spinning off its manufacturing and foundry operations (Intel Foundry) as a separate company. But while drastic changes may be necessary to ensure Intel's recovery, separating manufacturing from design wouldn't do the company much good. Let's break down the pros and cons.

Full Disclosure: While this article reflects my independent opinion as an analyst, Intel, along with AMD, Qualcomm, Arm, and Nvidia, is a client of my firm, Moor Insights & Strategy.

The potential benefit of a standalone Intel Foundry business

The only benefit of this spin-off for Intel is the opportunity for an independent manufacturing company to attract new customers who compete with Intel products – and who don't trust Intel's design department not to snoop around in their own chip designs.

AMD and Qualcomm top that list. I believe that if Intel's new 18A process node is as strong as Intel says – and defect density numbers look good so far – AMD may have to acquire it. At least that's what might happen if AMD doesn't get the secondary source of foundry capacity it needs from Samsung as TSMC continues to raise prices. Keep in mind that AMD really needs a secondary source of manufacturing capacity to meet its ambitions, which I recently discussed in the context of AMD's planned acquisition of ZT Systems.

Qualcomm, meanwhile, has a strong tendency to use the best available production technology wherever it is available, so Intel's more advanced 14A process node could fit – although it is also probably more than a year away from production. It's worth remembering that in such an arrangement with Samsung, Qualcomm is contracting for multiple production cycles. Qualcomm also sources its RF products from multiple suppliers. But Qualcomm has no reason to trust Intel if a fab spin-off is essentially a tracking stock; it would have to be a true spin-off into a fully independent foundry company.

I would add that Intel 14A is probably a more suitable node for mobile SoCs. 18A may be low power, but I'm not hearing anything positive about smartphone or IoT mobility. If Intel announces a smartphone SoC or IoT vendor as a customer, my opinion would change. There is promising news from Intel on 18A. Intel's AI PC client processor (Panther Lake) and its server processor (Clearwater Forest) are out of the factory and have been powered on and booted operating systems. In addition, Intel has announced that the first external customer tapeouts will be in the first half of 2025. Microsoft is one of them, and I suspect it's for Maia, its homegrown AI accelerator for data centers.

Keep in mind that any scenarios involving AMD, Qualcomm and other potential foundry customers beyond what has already been announced are hypothetical. could There may be a future where Intel Foundry wins more new and significant customers like this, but that future would look very different from the status quo. Today, Intel Foundry primarily produces Intel's own chips; there aren't many other customers, at least so far.

Why such a division no longer makes sense today

Sometimes there's a compelling reason — or at least a good argument — for splitting a large company into independent parts. Hewlett Packard's history provides numerous examples from different time periods, including the spin-off of Agilent in 1999 and the split of HP and HPE in 2015.

But that's not all. According to Wall Street calculations, Intel Foundry's current value is less than zero, meaning a spin-off would likely have to be done privately. How much value that would create is questionable. And there would still be all the challenges (and costs) of building a foundry customer base separate from Intel's parent company.

Could we imagine a scenario for going private? Sure, if we get even more hypothetical about the specific mix of debt, capital allocation, and investments from customers (Apple, Nvidia, and Broadcom?), that could happen. You still wouldn't be talking about a profitable, freestanding foundry company until 2027 or 2028. And that's before we get to the incredible upheaval to Intel's business model and internal culture of converting the design side of the company into a fabless company.

In short, if Intel spins off Intel Foundry before the design and foundry businesses are healthy, it will fail.

What is it really about?

While there are some bright spots, Intel is definitely in a serious crisis. It seems that the company is taking all possible steps to cut costs, including slowing down its factory expansion plans and canceling projects that are not able to become profitable sooner or later. It is also shifting resources from the 20A production node to 18A, which makes sense to me as a short- to medium-term move, as 18A seems to be on the right track. Employing more engineers should reduce design risks and make it even better.

If we want to be optimistic, production nodes 18A and 14A could turn out to be real winners, and I believe Intel could start to take market share from Nvidia in the data center AI space in 2025 and 2026. Intel should also see a boost as its Lunar Lake processors make their way into a large number of AI PCs in the coming quarters, and at some point – around 2026 – Intel should have a proper data center GPU (not the current Gaudi ASIC AI accelerator) to compete directly with Nvidia.

Given all of this, why consider a spin-off of Intel Foundry, which doesn't seem to offer any concrete benefit that would make the move worthwhile? I believe Intel leadership is doing this to stave off a potential coup, perhaps by recently resigned board member and semiconductor industry icon Lip-Bu Tan. By running through detailed scenarios with their bankers at Goldman Sachs and Morgan Stanley, Gelsinger and the rest of the board can tell investors, “We've done the analysis and this is the result. It's not a good idea to spin off Foundry right now. These are the conditions under which a spin-off would make sense.”

Call it preemptive due diligence. Intel has enough problems without another loss of investor confidence or a potential executive shakeup by Tan or others. So it's in Intel's interest to let investors know that it's exploring all options — even if the idea isn't viable. Maybe later, when the foundry and design business is booming, but not now. If you want to research a similar split, look no further than my alma mater, AMD, which spun off its design and manufacturing business as Globalfoundries. AMD had to do this to survive, and while Globalfoundries is very strong today, it was losing money back then and was privately funded by Mubadala Technology Investment Company. And so far, no such scenario is in sight for Intel.