Advanced Micro Devices Inc (AMD)

[bobs10]

Well I'm obviously no process engineer, but I've been through these transitions at least 4 or 5 times. The companies set up these milestones they call design rules, but it seems that the process is much more gradual and evolutionary.

With the current problems related to unsuitable low-k dielectrics and electron migration I see no reason for AMD to be on the bleeding edge of things. I would rather see AMD buying the 65nm equipment but not push the processes.

Given the advantages displayed on the 130nm process over INTCs best at 90nm and the expected advantages of moving to 90nm/SOI along with the advantages of the K8 design, these should be able to keep AMD well ahead of INTC for the foreseeable future.

So far we have limited information about how well AMD's 90nm process is going. While laptop chips may be going into product no one has done any tests on them. Given the problems everyone else has had with 90nm I'm reserving judgement until I see more information. Also, given the way these moves usually happen I would expect plenty of low lying fruit at the 90nm level for at least the next 6 months to a year.

All in all I see no reason for AMD to rush into 65nm, especially given AMD's checkered history of production problems with new design rules/steppings. It seems like the problems of 90nm will only be compounded at 65nm, and I haven't heard of any magic bullets. The best thing that could happen to AMD is for INTC to have the sort of conversion at 65nm they had at 90nm. And yes I realize that a lot of those problems were related to the Prescott design. Still, as far as I know, Prescott will be INTC's main product when INTC is trying to move to 65nm.