Hacker News Reader

Comment by bin_bash | Hacker News Reader

My understanding as an outsider to chipmaking is that Intel's model involved repurposing old fabs so they could make new ones out of the old. TSMC's model is to keep using the same fab for years and years.

Because the chip shortage is largely for older, cheaper silicon, Intel's model doesn't allow for building the chips that are actually needed right now.

View on HN

I think the other elephant in the room that not many are talking about is that we have reached a point where most people do not need the latest and great CPU anymore - many can still get by with something 5 years+ old. In the old days stagnation like this could have been a death sentence for Intel but modern times are much more forgiving if they can still play catchup eventually.

Isn't that just a sign that Intel hasn't managed to make any compelling products?

Back when Intel were really resting on their laurels it was hard to tell. Now we have M1/Ryzen you can reach some sort of conclusion depending on your view of those:

- doesn't unlock some major new application for computers, 99% just want Office 365, which ran just as well on Sandy Bridge -> whole industry stagnated, being a couple of years behind is fine

- (M1 especially) battery life/perf is improved in a way that will matter to many -> Intel specifically risk being left behind in sector that still has the potential for big changes

Intel can't start filling gaps in the market for car ECUs or whatever, but there has been a shortage of PC components too, people are buying whatever is available, so it's an ideal moment for Intel's down-cycle to be happening - sales numbers will be more about manufacturing capacity than the strength of the product.

Reply on news.ycombinator.com