Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: peter_luc who wrote (72520)	2/26/2002 6:53:55 PM
From: andreas_wonisch	Respond to of 275872

Peter, I think this doriangrey guy has lost any credibility he might have had a long time ago... Andreas

To: peter_luc who wrote (72520)	2/26/2002 7:16:52 PM
From: combjelly	Read Replies (1) \| Respond to of 275872

"First Hammer engineering samples out already 10 months ago?" I think doriangrey is just blowing smoke. It is possible that AMD actually did a Hammer tapeout at the end of 2000, but, assuming they did, they found something they didn't like. AMD talked during 2000 about a bulk Si, 0.18 micron ClawHammer that would drop in a SocketA but would run at a lower voltage than normal. There was an interview late in 2000 with Bob Minton (I think), a VP at AMD where he stated that by the end of that year they would have a Hammer tapeout. JC used to have it on his website, in fact, he has a link to it, but it's dead right now (JC, are you listening?). It's translated from Japanese, but it is pretty explicit about that. Anyway, soon after you would have expected first silicon, the Hammers abruptly got "aligned" with their 0.13 micron SOI process and there was no longer any mention of a SocketA version. So why did that happen? It could have been that SSE2 was deemed to be important, but that doesn't explain why they didn't go with SocketA, despite the low risk of using the existing infrastructure. I think the likely explanation is that the bandwidth limitations of EV6 was rearing it's ugly head and the ClawHammer wasn't going to have much, if any, performance advantage over the Palomino at a given clock rate. So it was decided to scrap the design and go with Plan B, i.e. and embedded memory controller. Since it is incredibly unlikely that AMD allowed any of the SocketA ClawHammers outside of the company, I suspect that doriangrey is full of it. Unless he works for, or has contacts in someplace like Compaq. I could see AMD showing some systems around Compaq especially, just to show what they are doing. But I wouldn't expect the benchmarks to be all that great. Now this would also mean that AMD does have a leg up on getting functional silicon quicker that one might otherwise expect. Under this scenario, they might be able to pull in the release date somewhat, but debugging and validating HT and the memory controller will still take some time. But with a little luck, Q4 systems are not out of the question... And maybe all of this explains why the 760MX didn't have HT as it was supposed to have. Since it wasn't going to be a vehicle for ClawHammer, it declined in importance.