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: Elmer who wrote (61465)	6/11/1999 12:48:00 PM
From: Tenchusatsu	Read Replies (2) \| Respond to of 1571996

Elmer, we're also awaiting actual benchmark scores, not just the "percentages". I still think Dirk Meyer has no reason to exaggerate, though, especially when the K7 is this close to release. I mean, what if independent web sites as well as OEMs have trouble trying to reproduce the "percentages" that Dirk Meyer tossed out on his foils? It will make Dirk look like a liar and AMD look like just more hype. That why it'd be pretty foolish to exaggerate the K7's advantages, especially now. Tenchusatsu

To: Elmer who wrote (61465)	6/11/1999 12:53:00 PM
From: Cirruslvr	Read Replies (1) \| Respond to of 1571996

Elmer - RE: "It is my guess that a Coppermine will match or slightly beat an equally clocked K7 in Integer performance. We will know before long." Yeah, that may end up happening! "Floating point is another matter. At this point I can't see any other way to interpret the K7 results. It appears to have a clear and definite lead that Coppermine will not be able to close, regardless of cache size." I see the paranoia is sloooowly fading away...

To: Elmer who wrote (61465)	6/11/1999 2:04:00 PM
From: Petz	Read Replies (1) \| Respond to of 1571996

Intel's low-end solution a no-go for gamers: (from c't magazine heise.de We thoroughly tested the first two [Camino chipset] boards that arrived at our office: the PW65-D from DFI and the 370SWD from Supermicro... This means the CPU also handles the tasks of a modem and a SoundBlaster card, only a codec component and a line interface is needed as additional hardware... As long as the CPU is not very busy it can easily deal with this. But considering modern networked games, for example the notorious Quake, when modem, audio, 3D graphics and CPU must deliver high performance simultaneously, the remaining frame rate might be fairly low. Our test boards were already equipped with an audio codec, to connect the modem we needed a so-called AC97 modem riser card. Even without softmodem it was quite evident that Whitney's UMA concept (like all UMA concepts) clearly affected the frame rate during 3D games despite the display cache. The cache could not be shut down for the test; according to Intel the performance of the 810 drops another 20 to 30 percent without the video cache. (cheaper versions of the Whitnet chipset do not include the display cache) Petz

To: Elmer who wrote (61465)	6/11/1999 5:53:00 PM
From: Kevin K. Spurway	Read Replies (1) \| Respond to of 1571996

Re: "How about real numbers to back up the SpecInt95 comparisons? A report on another website says the Xeon was 23.6 and the K7 was ~25 based on an eyeball extrapolation. The best published Xeon result @550MHz is actually 24.4 not 23.6. That would make the difference only 1/2 what AMD claimed at ~2.5%. It is my guess that a Coppermine will match or slightly beat an equally clocked K7 in Integer performance. We will know before long." Elmer, you're getting confused. The website you're referring to took a published Xeon performance number and added 5% to it! You're then comparing this to another published Xeon performance number and saying you're not impressed with the K7 result?! Are you SURE this is the argument you want to make? Kevin