This release puzzles me for two reasons. First is that it isn't needed at all, AMD has no part that is quad-ish in one socket, so where is the threat? The other problem? Power, this one is said to blow out the 120W limit and then some.
We are told that the NDA goes up on the QX6800 Monday, the samples floating everywhere now are 2.93GHz on a 1066FSB, basically matching the current X6800 clock for clock. There will be two of them though, so it should be a great gaming part.
The first snag is volume. Intel is basically warning OEMs to not expect many if any. Q2 production will be in the low thousands, not tens of thousands of parts.
The curious bit is that with a lack of competitors, why do this? If they don't price it above the QX6700, all they will do is drive the price on their volume parts down. What's to gain other than lost margins? I don't get it.
The next one is also puzzling in light of the market not needing the performance. We hear the parts are going to be closer to 150W than 120W initially. If you think about it, Kentsfield at 120W is 2x 65W 2.67GHz Conroes. 2.93 Conroes are 80W, so two of them minus a little is about 150W. The math works out.
The real QX6800s will come out in Q3 with a new stepping of the core Core number core numeral 2. This new stepping will drop power we are told 'substantially', and pave the way for volume QX6800 production. It also bodes well for a dual core speed bump, but that may not be in the marketing cards.
So in the end you are less than a week away from a hot and scarce part. This market may not care at all about the power draw but the scarcity is going to be more an issue. Then again, they only make so many Ferraris, and those get fairly bad gas mileage but no one complains. µ
Sign up for INQbot – a weekly roundup of the best from the INQ