I think Intel is heading into a world of hurt. They made too many mistakes over the years. And they relied way too much on their dominance in the PC processor market. They always stayed just one small step ahead of AMD. But in the semiconductor business one small step ahead isn’t enough because if you hit a small snag when moving to the next step (the next geometry, for example), your competitors can leapfrog you. And that’s exactly what happened to Intel, their competitors leapfrogged them. Then add in the fact that we may reach processor-agnostic systems at some point, and their lesser performance could kill them.
That doesn’t mean that Intel will disappear, but they will slowly wither and be relegated to lower priced, lower performance products. AND, whatever billions or tens of billions that government is handing them over the next 3-5 years will essentially be wasted. It isn’t only performance of the chips themselves, it is also the performance of the workforce. First of all, they will have a demoralized and disgruntled workforce (giant layoffs tend to do that). Second of all, they will have a hard time attracting big talent, who wants to devote 5-10 years of their primary career to a company that isn’t likely to pay off well for them? And third, they already have a reputation for lower performance and even shoddiness (including trying to paper over major security flaws) in the marketplace.
This doesn’t mean they will definitely fail. With the right management, with a lot of luck, and with billions of dollars from government, they might pull out of their tailspin. But it isn’t likely.
I generally agree, and the market certainly agrees. However, Intel now has both P-core and E-core servers on TSMC N3 generation technology. The Lunar lake mobile CPU is in production and manufactured on TSMC N3. That said, the bulk of Intel (and AMD for that matter) product is on older generations but at the leading edge Intel is now more than competitive.
The real problem Intel has is cost. If you look at Intel revenue share versus AMD it has been very stable for the last several years. AMD has steadily gained share in the server market, but offset by other declines. Based on the Intel CC, the Intel_3/4 process being manufactured in Ireland is currently not yielding as well as it did in the Oregon development fab, and this is putting a lot of pressure on margins. In addition, the anticipated high volume lunar lake product is not “cost optimized” so will actually degrade gross margins as well.
The Intel product portfolio looks pretty strong, but the financials look pretty terrible. This seems like a logical outcome of switching from a finance based CEO to a purely engineering based CEO. I think there are still a few more chapters to this story.
NVDA would I’m sure want to buy Intel and they are earning enough to do it. That could be a nightmare for AMD which would be a shame since they are the most inventive company in the semiconductor index.
AMD’s achievements make them a company I like to own; first (I think) to bring multicore, Opteron, 64 bit and sensibly priced to boot.
Intel is suffering, it could be a commercial match, and presumably would not need no-deal insurance for ARM that Nvidia had to pay last time. It depends more on the threat to Intel. If it’s seen as a rescue it by Nvidia the acquisition might be considered differently.