Some observations about Dell initiating mainstream AMD sales.
1) It’s been somewhere between six-to-nine months since Dell told its manufacturing partners to start getting ready to build some AMD machines for them. Knowing what we know now, it looks like the delay was largely, if not entirely, due to AMD needing to increase its capacity enough to fit Dell in and not shut out others. That capacity should be there by 4Q, and that’s just when most of the Dell order will end up being processed.
2) It’s hard to avoid under- or overestimating the impact of the initial Dell purchase. Assuming the two million CPUs ordered are for the next four months, and that 90% of it is meant for the Christmas quarter, this would be about 2.5% of estimated world market share Christmas quarter and a bit over 3% of Intel’s CPU revenues. So it’s hardly an Intel-killer.
That may not sound like a lot, but a shift of 2.5% is pretty significant when you consider that over the last ten years, marketshare has wiggled around in a range of just 10% even counting all the extremes, and usually has stayed within a 5% range.
3) We’re in a very odd sandwich situation, and will be for at least another six months. Intel has often had the best mainstream product line, sometimes it’s had the worst, but now it has both, with AMD sandwiched inbetween. Buying a Conroe over an X2 is a no-brainer, but buying an X2 over a Pentium D is a no-brainer, too.
What will be interesting is how Dell will arrange the layers of the sandwich, meaning just they’ll price the AMD systems compared to Intel. I suspect a lot of people are thinking that “of course” the AMD systems will cost less than the Intels, but I’m not so sure about that. Of course, the Conroes will be priced on the high side, and the Semprons will be down there with the Celerons, but I think the X2/Turion systems will be priced at or a bit above the D/Core Duo systems.