AMD, Arm, Nvidia. I’d argue the much bigger threat has been Arm and every Intel CEO (including Gelsinger) as utterly ignored that threat. Should’ve been an Arm licensee 15+ years ago. But no. The hubris is so strong.
I mean, go back far enough and they *sold off StrongARM* but I honestly thought that being a fab for anything you want including Arm was the smart way out of that. everyone out here forgetting how long and how much money it takes to build fabs
Arm has been inevitable for many many years. Too many in the industry (including Intel) ignored this for years. But folks aren’t laughing at us any more. They finally see how this is going to play out in the years to come
I wish Intel would just have embraced the Arm opportunity years ago. Its a tremendous synergy in terms of filling fab capacity and great for innovation
While I'd agree as it stands, Intel was on a path to fixing itself, which is not something I can say for the latter.
A bet on foundries is a risky one, but it did need time to play out. Pat had two options; fabless pure-design company (with huge capital raised from selling fabs) or high risk fab.
He definitely went for the riskier one, but it would have given them far more leverage and control if it played out. The capital risk, however, was way too much, I guess.
Let's see what the next CEO does, because the process is already in motion and pivoting would be riskier.
I don’t think Pat was a particularly strong leader but even if he had done all the right things, this was an uphill battle. Intel should’ve been an Arm licensee a decade plus ago when it had the chance, but hubris prevented it and Pat didn’t go hard enough to remake the company top down.
I suspect a major issue was that the hugely expensive foundry efforts drained too many resources from the product (design) side. Not being able to benefit from the AI wave was a huge miss, and AMD continued to steal particularly server market share.
AMD doesn't sell nearly as many AI accelerators as Nvidia, but they had a pedigree in the GPU segment and their MI300 sells in the billion dollar range. Intel's Gaudi has failed to gain traction, and they are now frantically working on the "Shores" chips. In other words, currently they have bupkis.
Hard to see what happens next. The appointment of the "CEO of Intel products" points to a refocus on the product side, but what will happen to the foundry? It depends largely on external customers to achieve the necessary scale.
It's at kinda the pivotal point rn. They're bleeding money and with the CHIPS act, they have some legal obligation to build some fabs at the very least.
Intel has missed multiple major waves under multiple CEOs. They missed mobile (and this was one that I count as the original sin. B/c by missing mobile and dismissing Arm, Intel missed multiple future waves), they missed graphics (multiple times), they missed lower nanometer, they missed AI.
Right, but most of that happened before Gelsinger. The biggest miss was the process technology that he tried to turn around. And AFAIK they are still on track with 18A in 2025, so it's not all bad. I suspect we'll soon learn why exactly Gelsinger was pushed out.
20A was scrapped. The BIG question is - how is 18A doing? 18A was supposed to bring node leadership. Panther Lake processor (based on Intel's cutting edge 18A) was supposed to be released 2H/2025. Gelsinger was fired, so I would be worried about 18A. Intel is probably toast if 18A fails.
then those people should remember that it only took one marketing dude to drive the company into the ground and going back to engineers is the only chance Intel has of surviving
I think unless they can hire an Allen Mulally type, they are cooked. And even then, I don’t think they have enough time. The time for the turnaround CEO was 2013, if we’re honest.
Comments
ok, this is a tad bit ominous for the entire industry.
Coming from almost anyone else, I'd have thought this was a parody headlone. 😭
Average wall street shortsightedness.
A bet on foundries is a risky one, but it did need time to play out. Pat had two options; fabless pure-design company (with huge capital raised from selling fabs) or high risk fab.
Let's see what the next CEO does, because the process is already in motion and pivoting would be riskier.
I totally agree with your point, design was the safer and more profitable bet for sure.