Once unthinkable break-up scenario might be on the table for Intel...
It used to be unthinkablebut Intel has, in its own words, run into economic problems, or at least a less than satisfactory situation. The company posted worse-than-expected revenue and a big loss in the last quarter, follow-up cuts and layoffs have already been announced. But it looks like this could lead to an even more significant historic turn. Instead of investing in chip technology, it could close its own fabs or end up broken up.
Intel is now in a precarious situation that largely stems from owning their own in-house own fabs. While most companies in the semiconductor industry like AMD or Nvidia are so-called “fabless” outfits and have their chips fabbed via outsourcing by contract manufacturers like TSMC (so called foundries), Intel has its own fabs. And these have long been seen as its strongest weapon enduring dominance over the silicon industry.
But while that was true a decade ago, since then Intel has fallen behind TSMC in both technology advancements and in the economies of scale, making its own fabs a hindrance instead of a trump card.
The current CEO, Patrick Gelsinger, returned to Intel with the goal of remedying this situation, which requires the company to make up for the technological gap in the quality and capability of its manufacturing processes, but also to compete with TSMC in another way – on the economies of scale of its manufacturing. It simply needs to produce more chips in more fabs than it does now in order to spread the extreme fixed costs of research and development (as well as the capital needed to build and set-up fabs) and thus being able to once again compete with TSMC on price. But to do this, Intel itself must move to a foundry production model and offer its fabs to outside clients whose contracts will supply the necessary revenue that Intel could not put together on its own.
The problem is that Intel doesn’t have such clients as of yet, and although it is trying to get them in the future, it doesn’t have the ability to tap that external manufacturing income for now. At the same time, it needs to build more production capacity, which requires huge investments. But it has become hard for the company to afford those, because the company has lost its previous high margins and profits and is facing negative investor sentiment. Intel made the mistake of not starting this plan and investment much earlier, when it was on its economic peak, and much of the money that could have been spent on these vital investments now has been spent on dividend payments and stock buybacks.
On top of that, Intel has now been hit with an embarassment at the product level when serious defects were discovered in Raptor Lake processors suffering failures and degrading after nearly two years in the market This can be a significant reputational blow, but also a financial burden due to the risk of serious RMA costs that these failures may trigger over time.
Tip: Confirmed: Raptor Lake CPUs are degrading, Intel preparing fix
Intel exploring sale of fabs or cancellation of some upcoming plants
Bloomberg has now come out with a report that Intel is already preparing for the possibility that the attempt to save its own fabs and manufacturing process development by transforming to a foundry model for external clients will not work. According to sources familiar with the situation who disclosed the matter to Bloomberg, Intel is consulting with advisors and analysts at investment banks Morgan Stanley and Goldman Sachs Group on how to deal with the situation. Such consultations are usually related to plans for acquisitions, but also for the sale of the company or parts of it.
What looms here are two scenarios in particular that Intel is discussing with advisors. The first would be a repeat of the situation AMD went through 15 years ago. That is, a process where Intel gets rid of the fabs because it doesn’t have the financial strength to pay for their operation and future development. This would turn Intel as a processor manufacturer into a fabless company comparable to AMD and Nvidia, which could even switch completely to chip manufacturing at TSMC over time, in theory.
It is likely that the remaining core Intel company would have a better chance of becoming profitable again in this mode. But the spun-off fabs, on the other hand, could be in a rough spot. The fabs likely generate the bulk of Intel’s losses, as they are apparently the area where Intel is not competitive and the costs of the entire operation exceed profits (in the last quarter the fabs had a loss of $2.8 billion).
The separation of the fabs would probably not mean their immediate shutdown. Intel would probably try to sell them or spin them off into a separate company, similar to how AMD’s fabs became GlobalFoundries. But Intel would need a strategic partner to buy or enter the foundry business and supply the funding that Intel is now having trouble securing on its own. One option here is probably an IPO, i.e. the creation of a new public company around these fabs. By selling an emission of its shares, the construction of the fabs and the development of the technology could be financed.
The end for comeback ambitions?
This move would, of course, devalue the original Intel and its shares. But most importantly, it would limit Intel’s ability to regain that prominent vertically integrated position it had in the past – the company would have to get used to no longer being a special case among chip companies, but just the same kind of fabless company like Qualcomm or AMD are. This could lead to a long-term decline in market share and, with it, sales, so Intel could gradually shrink and thus worsen its starting position against competitors. This would effectively thwart Patrick Gelsinger’s existing plan to regain a stronger position (by reinvigorating the fab base of the company).
Even a fabless company can rebound to unprecedented revenue growth like Nvidia is showing now, but that would also require Intel to find some new rapidly expanding (bubble?) market where it would have a dominant position, something the company hasn’t managed to do in a long time. Besides, even the fab spin-off would not be guaranteed to succeed, they could end up as a second-rate player in the foundry business, as it happened to GlobalFoundries created from AMD fabs. Independent Intel fabs would probably only have reduced chance of catching up with TSMC, and it would probably further consolidate TSMC’s position as the chipmaking hegemon.
According to Bloomberg, the above scenario is more of a last resort for Intel. Before that, as an alternative, the company will probably try to downsize its plans to build new fabs, which is also one of the things that investment bank advisers are suggesting. In this case, they would stick with the plan to create a foundry business competing with TSMC, but it would be with fewer fabs and investments. Thus, from a worse position.
Intel is reportedly analyzing which of the currently planned or already implemented projects for new fabs and production lines could be cancelled to save money, or at least postponed or suspended. This is said to be a concern, for example, in Germany, where Intel wanted to build its first chip manufacturing plant on the European continent (Intel already operates in the EU, however, with a fab in Leixlip, Ireland).
According to Bloomberg, the board will also discuss these options in September, but the exploration and consideration of these scenarios is still at an early stage. So any radical decisions are unlikely to come soon, perhaps not until next year. In any case, this is a big warning that the company’s position may be quite uncertain and that, although for many years the company seemed unshakable, accumulated internal problems may suddenly burst out and put Intel into a full crisis.
Source: Bloomberg, via: techPowerUp
English translation and edit by Jozef Dudáš
⠀
