Intel and the Last Foundry Standing Theory

Coming out of the Intel Developer Forum (IDF), I thought I would share more on what I believe is Intel’s long-term thesis about themselves. Quite simply, I believe they feel they will be the last foundry standing with their leading edge process technology. If you follow this as closely I do, you will note Intel and other foundries like TSMC and Samsung are in a race to get to the next process node. Today, these foundries manufacture at many different nodes but 14 nanometer and 16 nanometer are the leading edge processes of today. With each jump to new process nodes (10nm being next and 7nm after that), designers can pack more processing power onto a single chip while still keeping a low power profile. The industry calls this “performance per watt” and the amount of performance increases with each new process while still maintaining a lower power voltage.

Companies who own factories that make chipsets care about this from a competition standpoint. By getting to the next process node, it makes your factory more attractive to potential customers. It is also important to point out that getting to the next process node is not only expensive but extremely difficult. We are talking about advanced quantum physics and, at this point in time, the industry is not sure of the science to get their foundries past 10nm. 7nm remains mostly a theoretical goal, with some technologies in the works being tested out which may make it possible. The two biggest things impacting this, and Moore’s Law in general, is economics and science. It is getting more costly and more difficult scientifically to keep pursuing leading-edge process technology.

This is what makes part of this analysis so fascinating from an economics perspective. If it continues to get more costly to pursue leading-edge process technology then, unless you have customers who will pay a premium for those advancements, it is not worth pursuing. There are only a few name companies who will pay for this and one of them is Apple but Intel is also in the mix since they still sell somewhere north of 350 million chipsets annually when you add servers. While costly, Intel is still well positioned to pursue leading-edge process technology — perhaps Samsung is as well. For TSMC, it is a little more tricky as their largest customer willing to pay and utilize the latest node and ship SoCs in volume at that node is Apple. Should they lose Apple, getting beyond 10nm would be very tricky economically.

I emphasize the difficulty getting beyond 10nm will present. There is a saying of many industry vets: “The harder it is for Intel, the better it is for Intel.” This is ultimately the root of what makes up my last foundry standing theory. It may genuinely come down to them and Samsung and, while I certainly will not write off Samsung, you certainly can’t write off Intel either.

Now, let’s address the “Then what?” scenario. Let’s assume this theory is correct and Intel is the only one to get to 7nm or less. Either way, they are the only one who can manufacture at that node. Then things get interesting because time wise, we are talking about the 2022-2025 time frame, which is where we have to believe, as an industry, the semiconductor segment is shipping more semiconductors than they are today. Including chipsets and digital sensors, over a trillion semiconductors are shipped annually in 2016 and that number will grow as we connect more things. Intel has a lot of capacity in their foundries but they don’t have enough to meet the needs of the global semiconductor industry on their own existing capacity today. So they either need to acquire more capacity or license their process technology to other foundries like Samsung, Global Foundries, TSMC, and accrue revenue on the chipsets they make for others on Intel’s process. In this scenario, Intel gets licensing revenue from licensees and high margins from their own first party manufacturing. Not a bad business at the end of the day if it pans out.

There is, of course, the scenario where Intel is wrong and others do keep pace with them on process technology. Part of their manufacturing business plan to compete in manufacturing ARM chips for others is a segment of this backup plan I think. While we can argue they should have done this sooner, if we look back in 10 years and others have kept pace or beat them, then it could be said their lack of willingness to compete for ARM manufacturing earlier and more vigorously was financially irresponsible for shareholders.

We will see if and how this pans out. For me, it is a central theory for how I analyze Intel and others in this space. At the end of the day, right or wrong, hopefully, you find this theory to be interesting at the very least.

Published by

Ben Bajarin

Ben Bajarin is a Principal Analyst and the head of primary research at Creative Strategies, Inc - An industry analysis, market intelligence and research firm located in Silicon Valley. His primary focus is consumer technology and market trend research and he is responsible for studying over 30 countries. Full Bio

5 thoughts on “Intel and the Last Foundry Standing Theory”

  1. Two things. Why would Apple want to kill the only competition that exists for Intel by giving only Intel their business for the next generation process? It would seem to me they are better off helping TSMC develop their own 7 nm process to keep their future foundry options open. If TSMC and Samsung stop developing smaller nodes, Intel can charge Apple whatever they want in the future.

    Second, if the 7 nm node doesn’t get here until 2022 or later, wouldn’t that give TSMC and Samsung plenty of time to catch up? Wait for Intel to prove out the technology, buy the same equipment and work out how to get to 7 nm after Intel proves it is possible. I would think this would reduce the R&D cost considerably. Since Moore’s law is slowing down, isn’t it likely that even foundries with lesser R&D budgets than Intel can catch up eventually because Intel isn’t racing away like they used to. If 7 nm is still mostly theoretical, 5 nm is going to be science fiction at this point.

      1. “There are rumors about Apple in talks with AMD right now as well (but for x86 not ARM).”

        That rumor has been going on for years. Just mere speculation.

        1. Yes. But given Apple’s current seeming indifference to updating any of their CPUs, I’ve given it more credence than usual. But it is just speculation. I haven’t seen anything that implies a leak.

Leave a Reply

Your email address will not be published. Required fields are marked *