It’s surely not intentional, but Intel looks like they’re giving AMD a chance to get back on its feet with its Bloomfield positioning.
XbitLabs has seen a copy of Intel’s projected product marketshare roadmap, and they’re just not going to make many Bloomfields, ever. Let them explain:
The share of Intel Core i7 chips is expected to be below 1% of all microprocessors that Intel will ship in Q4 2008 . . . . [It] will hardly rise above 3% even in Q2 2009. . . . [E]ven in the third quarter of next year . . . the share of Core i7-based platforms is projected to be below 5% of all CPUs sold by Intel.
If you’ve doubted Intel was out to create a luxury brand with Bloomfields, this should remove them. This is not a typical new-generation product ramp for Intel. Historically, after three quarters, the proportion of new generation processors is more like 40%, not 5%. The only Intel product ramp we’ve seen that resembles this pattern was that for quadcores.
So how does this help AMD? It helps AMD because its 45nm chips will compete in the real world not against Nehalems, but Penryns. A real-life matchup between Denebs and Penryns ought to be a lot more competitive than we’ve seen in quite some time, and Intel will have nothing to trump that in the mass market until 3Q 2009.
Of course, AMD has to price them right, which in AMDland means they have to be able to make sizable quantities of them. We’ll find that out when we get initial pricing. If they can’t make many, they’ll price them sky high and milk the most out of meager production. But if they can, we’ll see Penryn-like pricing, and if that’s the case, I think AMD will have its first winner in a long time on its hands, both in the OEM market and, if they can overclock decently, even in the enthusiast market. Not like AMD is going to take over the world or anything, but when you have the enemy flat on his back, you shouldn’t let him get back on his feet just to make a relative handful of extra dollars. nVidia made that mistake, and Intel looks like they’re going to repeat it.