It seems that Intel is having a few problems with its 45nm process, and it looks like the problem with getting them fast.
Why do I say that? Well, according to this roadmap, 45nm chips not supposed to show up until 2008.
Is that really a delay? Well, given that for-sure Penryn compatible motherboards are supposed to show up late 2Q/early 3Q, and it was reported in September that the quad-core Yorkfields were supposed to show up 3Q, why, yes, that is a delay.
OK, those are pretty strong indicators of a delay, but why do I think speed is the problem?
Well, the initial models won’t be any faster than what will be the current 65nm models. Indeed, these are such a not-big-deal that Intel isn’t even going to change the model number for most of them, just like AMD didn’t on its current, less-than-awesome, 65nm chips. Nor is there a top speed on the top-end product at all.
All this isn’t a smoking gun, at least not yet, and its understandable Intel could have some of the same growing pains with hafnium and metal gates that AMD/IBM have had with SOI, but a delay is a delay, and this is a very bad time for Intel to have it.
Why? Well, in the middle of the year, AMD is going to come back with its response to C2D. Odds are, they’ll be able to catch up or at least get fairly close to C2Ds.
Now if Intel could just pop out these 45nm chips and jump right back ahead again, spoiling AMD’s show, that would really make AMD look like an also-ran that just got lucky for a while, and that would really, really bad to a lot of observers, both technical and financial. If you think AMD’s been getting bad press lately, what you’ve seen is nothing compared to what would happen if that happened.
However, with this delay, AMD can catch-up, get some good press, and if the initial 45nms aren’t any better than the 65s, that’s much better than trying to get up from your knees only to get kicked right back down again. Even if the 45s do end up pushing AMD back, waiting six months for AMD to come with 45s is a lot better than a year.
Yes, this is all public relations stuff, but much of the AMD surge is built on perception, and the perception of a comeback against Goliath will certainly help a lot more than Goliath delivering a one-two punch.