To: Dan3 who wrote (148630 ) 11/17/2001 12:48:08 AM From: wanna_bmw Read Replies (1) | Respond to of 186894 Dan, Re: "Next quarter, Intel will bury AMD, eh?" On the contrary, I am skeptical. I don't think that Northwood will have a major advantage over Athlon, but rather that the two will be more evenly matched. I believe that AMD's current "conservative" model numbers will start looking more "agressive" instead, but not so much that AMD is in danger of losing the PR battle entirely. I think there will be brief periods with Intel ahead with Northwood, and then AMD will get even again with Barton. Both of these are great micro-architectures, and what can I say...? I have faith in both of them. What I don't have faith in is AMD marketing being able to maintain their ASPs after Intel fixes their supply issue. I think that more fab capacity and the transition to .13u will allow Intel to produce more than enough Pentium 4 chips to hold demand. AMD will try to increase production at Dresden, but will likely be pressured in continuing their launch of higher speed bin, but lower yield parts. This will make it difficult for them to maintain market share, so they may lose another point or two over the next six months. In the months before Hammer, I think it will be the hardest on AMD, and I think that the actual Hammer launch will probably be delayed once more to January or February 2003. They will then get a boost in performance, probably outperforming the Northwood at the time, which I'm guessing will be >= 3.0GHz. By this time, I agree with some of the predictions of Sledgehammer being cancelled, and Clawhammer being the UP/DP server solution for AMD. I see it being a great performer, like the Athlon MP, but I see Hyperthreaded Prestonia chips having the edge in performance (not a lot, but enough). Desktop Hammer will probably have the lead until the next Intel launch, but visibility is cloudy at this point. If all Intel does is a dumb shrink of Northwood on .09u, they may get a megahertz boost to outrun Hammer temporarily, but I can't imagine this being a long term solution. Therefore, some micro-architectural enhancements may be necessary. Still, that's a long way off, and even first tapeout of a 2003 product won't occur until next year. Therefore, Intel still has time to make some performance enhancements, if they haven't already. Do you have a different analysis, Dan3? wbmw P.S. About RMBS, I think you should go for it, but timing is everything. See if it tops off some more first before shorting it.