Intel (NASDAQ:INTC) is breaking up. This will have important implications for INTC stock holders.
Not literally, but figuratively. CEO Pat Gelsinger said on February 22 that the foundry business, which manufactures chips for Intel and others, will now be run separately from the design business, home of the Intel x86 franchise and other light classics.
Best of all, the first big customer for the new foundry business will be Microsoft (NASDAQ:MSFT). It’s putting its ambitions for Artificial Intelligence dominance into Intel’s coming 18A chipmaking process.
How did we get here, and should you buy or sell Intel on this news?
Intel Software Failures
Computer hardware has been software for decades now, and Intel has been failing at it.
That is, the design of a chip, written in software, defining what it does, is more valuable than the chip itself. Intel lost ground in mobile, then graphics, and finally central processing chips, over the 2010s by forgetting that design and manufacturing are separate businesses.
While Intel hardware floundered, Microsoft software prospered. It’s now worth more than 10 times what Intel is worth, even when Intel’s debt of $42 billion is factored in. This lead has accelerated over the last five years. Microsoft stock is up 269% since early 2019, while INTC stock is down 18%.
The last big Intel franchise to go, of course, was in graphics processing. Nvidia (NASDAQ:NVDA) now dominates it as IBM (NYSE:IBM) did mainframes in the 1960s. As I have stated repeatedly this is also thanks to software, in this case Nvidia’ CUDA development framework, the Windows of AI.
The Microsoft-Intel alliance is aimed at taking Nvidia down a peg. Microsoft will customize chips for its own software, tossing OpenAI into the alliance. Intel will focus on the hardware, with $15 billion in orders depending on it.
Intel Subsidized
None of this would be possible without the Biden Administration. It promised subsidies for semiconductor manufacturing and delivered them through the CHIPS and Science Act. Intel is collecting enough to re-shore manufacturing to Arizona and Ohio. This reverses a trend where chipmakers favored lower-cost East Asia foundries. This is crucial for INTC stock holders to understand.
Ohio isn’t yet ready. Intel’s next two chips will be made by Taiwan Semiconductor (NYSE:TSM). But its foundry plans see it making chips with lines just 1.4 microns apart (that’s 14 Angstroms) by the end of the decade. This will mean mastering and even expanding on TSMC’s own Extreme UltraViolet (EUV) technology, which has brought circuit lines within 3 microns. (By way of comparison, the 760 units in a fly’s eyes are each 5 microns across.)
This is the big risk in Intel’s Foundry announcement. Can its engineers deliver on the product roadmap it has set out, for itself, for Microsoft, and for other customers (including, theoretically, Nvidia)? If it can, America and Europe will dominate the next era of chipmaking. If it can’t, China will.
The Bottom Line
Tech trends can come and go in Internet Time.
Intel’s turnaround is already 3 years old, and it will take years more to prove out.
That’s why analysts are telling investors not to pull the trigger on Intel (as I have) just yet. Some even have a sell rating on the stock.
I find that idea unpatriotic. Intel’s new foundries are the Apollo program of this decade. Success promises a nearly unlimited supply of high-throughput chips for AI applications. It will cut the costs of AI not just by 10 times, but by 100 times.
If you have a long-term view of the market and AI, it’s vital that you’re in on that. When you get in is a question for market technicians. Just make sure you’re there when the new chips start flowing.
As of this writing, Dana Blankenhorn had LONG positions in INTC, MSFT, TSM, and NVDA. The opinions expressed in this article are those of the writer, subject to the InvestorPlace.com Publishing Guidelines.
