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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Dan3 who wrote (91773)	2/6/2000 4:37:00 AM
From: Process Boy	Respond to of 1577232

Dan3 - <. I posted here some months ago speculation that the cache added to coppermine(that results in it being competitive with Athlon) may have been "borrowed" from the Willamette design. Which would mean that we've already seen a nice chunk of Willamette's performance improvement.> I would be extremely surprised if this were true. I think Willamette is a different animal. Will probably have to wait until technical details of Willamette are released before we know for sure. <but now you are pointing out that the coppermine cache doesn't make much sense as a match to the PIII core that it's been attached to> Dan, it does make sense. It provides 2X bandwidth over .25 Xeons, and 4X bandwidth over PIII Katmai's with 512 off chip. PB

To: Dan3 who wrote (91773)	2/6/2000 4:55:00 AM
From: Tenchusatsu	Read Replies (1) \| Respond to of 1577232

Dan, <now you are pointing out that the coppermine cache doesn't make much sense as a match to the PIII core that it's been attached to - are you sure that that cache wasn't taken from Willamette (or Itanium)?> First of all, I never said that the cache doesn't make sense, nor did I say that it's not a good match for the P6 core. In fact, the opposite is true; we see how much of a performance improvement Coppermine's cache provides. What I am saying is that the cache could have provided even more bandwidth if the P6 architecture allowed for it. Second, I doubt the cache was taken from Willamette. Two very different architectures. There will be just too many peculiarities between the Willamette cache interface and the P6 cache interface to allow for such a transfer. And third, Itanium will start off with an off-chip cache, perhaps Intel's last processor with off-chip cache. It will be custom full-speed cache, just like the current Xeon. I don't think there are plans to add an on-die cache to Merced. Tenchusatsu