Bloomfield looks to be the beginning of a long-term luxury CPU line.
The more I think about Bloomfield, the more it looks like a radical break for Intel, and I’m not talking about performance.
What Intel is doing with Bloomfield is essentially expanding its Extreme processors into an Extreme processor line. Bloomfields will have their own socket, incompatible with any eventual lower-end Nehalem. It will initially need an X58 motherboard, and don’t hold your breath expecting more reasonably priced Nehalem chipsets, needing three channels of DDR3 RAM.
This is going to be wonderful for those premium and luxury PC lines. Configure systems using the mainstream and premium lines, and you’ll find yourself saying, “Why should I pay a good deal more for the same components in the premium line?” Bloomfields will allow the upper-end lines to distinguish themselves, indeed, this is probably the main reason why they exist.
The entry price for the CPUs will be around $284, $562 and $999 for 2.66, 2.93 and 3.20 GHz, and those price points will remain for the next twelve months. In Q2, maybe Intel introduces a 3.46GHz and bumps the 3.2GHz and 2.93GHz down a price notch, but that’s hardly a “price cut” for overclockers. According to the Intel roadmaps, the 2.66GHz is supposed to drop off, but even if it doesn’t, the price on that will probably just drop to $266 or around $240.
x58 mobos will get cheaper over the course of the year, but they won’t get cheap. DDR3 will probably drop more in price, but that won’t get cheap (especially by today’s DDR2 standards) either.
Yes, you’ll pay less for these things if you wait, the problem is by the time the savings become significant, it will be time for Lynnfield. The most cost-effective Bloomfield box may be the one bought early with relatively little DDR3 to start, then more later.
So Intel is providing some pretty strong incentives, if negative ones, to buy right away. Either that, or wait a year for what may end up being unoverclockable chips, and if you don’t like that, Intel says, “Let them eat Penryns.”
You may not like this very much, but the real significance of this move seems to be that this luxury line idea is a long-term strategy.
Mainstream Nehalems are supposed to show up in 3Q 2009. A quarter or two later, the 32nm shrink, Westmere will show up, and that will replace Bloomfield
(Will these Westmeres work in Bloomfield sockets? Good question, we don’t know, would be good to find out, though having six cores rather than four may mean the answer is no.
Can there be much doubt that when mainstream Westmeres are ready, the processors based on the Sandy Bridge architecture will be the next luxury CPU line?
I suspect the average person reading this doesn’t like this at all, may even be angry, but point the anger the right way. Getting mad at Intel for maximizing its profits is like getting mad at an attack-trained pit bull attacking after you take the leash off. This is what Intel does when AMD isn’t around to constrain it, slow the pace, maximize the profits (and AMD would do exactly the same thing if the shoe were on the other foot).
This is the world without AMD, and so long as AMD doesn’t have CPU that can compete on the higher-end, that’s the way it’s going to be.