The fabs they own now aren’t competitive at the high end. Though I like to point out that this is myopic way of viewing the industry. Not every microcontroller needs TSMC 3nm; they’re perfectly happy running them off for tens of pennies per unit on ancient nodes, and those chips go into every little piece of electronics that isn’t thought of as a “real” computer. A huge portion of the pandemic supply chain issues (the ones that were actually valid and not just an excuse to raise prices) came from these little chips.
With that caveat in mind, the high end is still important.
The calls of “they took government money and then laid off employees” are off base. That government money is going into specific fab projects. Has nothing to do with the rest of the company. Unless there’s some kind of siphoning of funds into the rest of the company (admittedly totally possible, but it needs specific evidence), that’s not particularly relevant to Intel’s situation.
The company should probably reorganize as strictly a fab and stop making their own stuff. AMD did very well for themselves by ditching their fabs into a separate company–Global Foundries–and then buying manufacturing capabilities from the best fab they could afford. Intel could go the opposite way, with AMD becoming one of their customers.
My computer has a 45 nm CPU and is (mostly) fine. The only reason I’m upgrading is because it’s just barely old enough to not have UEFI which means Windows 11 can’t be installed which means in a little over a year it’ll be EOL.
Intel fabs can’t catch up, that’s not a recipe for success for high end chips. And that would mean they would face competition which means they’ll be dead.
Sure they can. Intel is already gearing up 20A and 18A nodes, and those should leapfrog TSMC. Not that TSMC is staying still, either, but Intel will likely have the world’s best density for a while.
Meeh I’ve been hearing this in investor meetings for too long. If it was half true they wouldn’t be so behind in their current node in stores.
