Charles - <<I "changed my mind" about AMD vs. Cu several months ago, becoming agnostic.>
Interesting! How did that happen?>
I actually posted this several months ago on one thread or another. I believe it was to kash, specifically regarding Copper contamination issues. At the time, I believe my post went something like, "I've studied up on the issues, and I'm no longer as concerned as I was."
I still have reservations about the maturity of the tool sets, but over time this will ultimately be resolved one way or another, just a matter of how painful this is going to be, or not. Kind of like speculating on Y2K. As far as I know, no one is yet shipping a high volume Cu logic process, although IBM and MOT are probably starting to achieve something close to what the industry would consider high volume.
<But, I don't buy the (lack of) performance benefit part. Some fairly reliable folks have suggested over time that moving to Cu at the current generation would give ~10% gain. (some more, some less to be sure but the 10% is a good consensus number).>
Intel's number, for it's methodologies, did not approach this. There is speculation that with tighter DR's, maybe the benefit increases. Thus, AMD may well receive more benefit than Intel. This is just a WAG, and not totally mine.
PB
|
