Most people reading this have by now seen Intel’s(NASDAQ:INTC) Q4 2022 earnings. The consensus is, they were horrible, and the guidance for Q1 was even worse. I agree.
The stock didn’t really go down as much as might have been expected based on how bad the earnings were, and that’s because the fundamentals the company is based on look better now than before the earnings; there was a lot of good news.
The bad news we kind of knew about with Microsoft’s (MSFT) earnings, and Intel always saying they didn’t see anything good in the near future with respect to the macro-economic environment. Even with that, the earnings were remarkably poor, so that’s definitely something to consider. But, some of it was also burning through inventory levels, and Intel shipping fewer parts than were being consumed. They mentioned this will be particularly evident in Q1. But, these are really short term problems, and don’t factor in on whether I buy the stock or not.
The good news I mentioned does. There’s a lot, but oddly wasn’t really accentuated much in the earnings by Intel.
Perhaps the most promising is, Intel’s server group had higher revenue than was expected. More than this, Sapphire Rapids is now out, and the company believes it will stop, but not reverse, the market share losses in this segment. Intel is ramping up production, and believes it will ship over 1 million parts the first half of this year, which isn’t really that much. I will admit to being a little skeptical of it maintaining market share, because it does not seem to be as good at AMD’s (AMD) parts in broad usage scenarios. Certainly it closes the gap, and has specific scenarios where it’s extremely effective, but will that stop market share losses? I’m not sure. Intel will also release Emerald Rapids later this year, but this is based on the same Intel 7 fabrication technology, and will only be a relatively minor improvement over Sapphire Rapids. Granite Rapids will be a new architecture, based on a new node, but will not be out until 2024. So, my own belief is, Intel will significantly slow the market share losses in 2023 for their server processors, but it’s hard, for me, to see them stopping it entirely. Of course, I could be wrong. But even slowing the losses will be something better than what’s been going on with Ice Lake being their leading product. Reversing market share losses almost certainly won’t happen until 2024 at the soonest, and it’s far from certain it will happen even then.
Another little tidbit Intel slipped in was they believed they took market share in their CCG group. I find this easy to believe, despite the low revenue. Intel has a superior product line, overall, but not overly so, so that’s a positive. More than that, AMD had to significantly reduce prices of their processors almost immediately after releasing them, which in addition to being pretty embarrassing, shows they were not selling particularly well against Intel. . Also, Intel has been pretty open about market share losses, when they occur, so I tend to trust them at this point. And then we have a previous market share report from Mercury indicating AMD got crushed, but this isn’t recent anymore so isn’t as definitive as it once was. Even so, I believe Intel is continuing to take share from AMD in desktop and mobile markets.
Perhaps the best news is the roadmap seems to be completely intact, and we’re getting a bit more color on it. Meteor Lake is rolling out in 2H 2023, and is based on Intel’s 4 node. After all the delays with so many products, and nodes, Intel’s ability to get this out this year and in quantity is a very favorable piece of news. It’s still not clear if this will be a mobile only part, especially initially, or if they will bring this to the desktop. Keeping in mind that mobile is around 80% of the market, and Raptor Lake is currently the leader in performance on the desktop, it is very possible Intel will release this only for mobile. The other reason I believe this is, what is Intel going to do with their lines if they move everything over to Meteor Lake? I say that because this is a multiple chiplet design, and only part of it will be made at Intel. The GPU will not be, and likely I/O will not be. With only the CPU made there, and it being on a new node (which will almost certainly make it a lot smaller), these chips will not use a lot of fabrication capacity. And when Intel doesn’t utilize their fab space efficiently, margins drop pretty substantially. So, it kind of makes sense to keep making chips on Intel 7 at the same time, and the move it toward “3” when their next generation processors are out, and also when they fab more chips for other companies.
Speaking of which, Intel’s fabs showed increased revenue to the tune of 30% year over year, and now show bookings for over $4 billion, with seven of the top 10 companies as future customers. While it’s not a big part of Intel’s business right now, IFS figures extremely prominently in the future of the company, with an enormous TAM that could catapult the company’s market capitalization to many multiples of its current value. While selling more processors can certainly help the company’s stock price, the foundry has the greatest potential to increase the stock price. In many ways, the future of the company depends far more on their foundries, than their design. Not the least of which a strong node helps their designs as well as wins new customers to make chips for.
Which brings up the last piece of good news. Intel again stated their nodes are all on or ahead of schedule. This and the buy in from potential customers I mentioned above, is more salient, by magnitudes, than a quarter or two of soft demand based on customer high inventory levels, and a soft macro-economic environment.
Short term, it doesn’t seem like there’s going to be much movement, because a lot of the catalysts for growth are still a bit down the road. But they are getting closer all the time, and they aren’t being pushed back like they have been. When they hit, people will be wondering why they didn’t buy sooner. Beats me. I’m all in.
Be the first to comment