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: Scumbria who wrote (52873)	3/18/1999 6:56:00 PM
From: Cirruslvr	Read Replies (1) \| Respond to of 1578704

Summary of AMD at CeBit "AMD AMD showed several K6-2 and K6-III machines on the floor and had a special exhibit in a caboose on top of their booth where they ran Winbench on equally configured PIII/500 against a K6-III/450, where the K6-III visibly led the race. Nothing really new but nicely demonstrated and the polish computer traders I slipped in with were duely impressed. The cherry on top of course was the K7 system that ran on the table beside that, it was clocked at 600 MHz and showed first that it would run the Winbench itself without stability problems and visibly a good deal faster than the other two machines. As a second sample a demo of Viper Racing ran on this K7, which worked together with a Voodoo3-2000 for display, which worked just fine as well, no frame rate display though. Finally AMD gave me some information on the K7 and board used, which I took a picture of to be seen here tomorrow. The K7 will initially ship in the second quarter, at speeds in excess of 500 MHz and with 512 kB to 8 MB of L2 backside cache, clocked from 1/3 to full CPU speed, depending on your budget and needs. The demo K7 ran on a reference design board with their own AMD chipset." fullon3d.com The article kind of hints that the AMD display was public, but doesn't say outright. I emailed the author of the article to find out.

To: Scumbria who wrote (52873)	3/18/1999 7:06:00 PM
From: Elmer	Read Replies (1) \| Respond to of 1578704

Re: "Going back to the beginning of development. I view K7 as being about 18-24 months old vs. 7 or 8 years on the PPro. Most of the speed improvements usually come in years 3-4 of a processor's life cycle." Not likely. That would mean the PPro was in development for 4 1/2 years before introduction. EP

To: Scumbria who wrote (52873)	3/18/1999 7:25:00 PM
From: kash johal	Read Replies (2) \| Respond to of 1578704

Scumbria, Re: K7 vs Xeon's I will throw out this question to you and onayone else who might know. In comparing the price/performance ratio's of the K7 vs Xeon. Let's assume that K7 is 30-40% faster than equivalently clocked Xeon. The AMD's will use interleaved memory as I understand it. Does that mean for a server that one needs 2x this memory for K7. It would seem to me that for a server with 1Gb of RAM that if it needs 2x that memory for K7. This would eliminate any K7 price/performance advantage would it not? I know folks who run up to 4Gb of main memory. So it would be much cheaper for somebody to get dual or even quad CPU's than dual memory for memory intensive servers. As regular DRAM moves to 133Mhz speeds the Xeon servers will gain performance with respect to a 200mhz K7 bus unless K7's will run at 266Mhz at that time. And once camino comes out with Rambus RDRAM then Xeon performance will gain again. So in Q3 when the battle starts it would appear to me that the total system costs may still be in the Xeons favor. Not a particularly appetizing scenario frankly. Regards, Kash Johal