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: Daniel Schuh who wrote (10009)	9/23/2000 12:38:25 PM
From: andreas_wonisch	Read Replies (1) \| Respond to of 275872

Dan, Re: I put more stock in them than the other leaked benchmarks, because spec isn't something a random hardware site can run. With the c't comments I think we can be fairly certain that the leaked numbers are the "official" ones from Intel. Unless Intel is downplaying here (which I doubt) we have a pretty good indication of what P4 performance on heavily optimized "application" will be. C't mentioned its own Mandelbrot benchmark (fractal rendering) which was very slow according to them -- that means legacy FP performance must be really bad. They also mentioned the Swim benchmark from the SPECfp2000 suite which "explodes" on the P4 (their words). Do you know if this some kind of bandwidth benchmark? BTW, I just reread the c't article and noticed that according to ITspace AMD reached 23% market share in Germany in the first quarter. Andreas

To: Daniel Schuh who wrote (10009)	9/24/2000 2:09:50 AM
From: jcholewa	Read Replies (1) \| Respond to of 275872

> the numbers indicating pretty good FP but mediocre integer performance were also consistent with what we know about > Willy. Of course, the good FP performance is dependent on compiling for SSE2; Intel has given warnings about x87 FP > being deprecated, so it won't exactly be a surprise if "legacy" FP performance is problematic too. Legacy FP is > here defined as not optimized for SSE2, and thus runnable on a few hundred million PCs, vs. maybe runnable on a few > hundred thousand PCs by the end of the year for SSE2 optimized apps. Dan, another thing to keep in mind is that specfp is very memory intensive. The extra memory/chipset bandwidth, I imagine, would be of a tremendous benefit to this benchmark.

To: Daniel Schuh who wrote (10009)	9/24/2000 12:51:12 PM
From: that_crazy_doug	Read Replies (1) \| Respond to of 275872

<< Legacy FP is here defined as not optimized for SSE2, and thus runnable on a few hundred million PCs, vs. maybe runnable on a few hundred thousand PCs by the end of the year for SSE2 optimized apps. >> Nvidia will almost certainly have SSE2 optimization in their drivers, and that will improve all old applications too. How much improvement? Well we don't know how good/bad it is now, so it's hard to say, but it might be enough to make it clock for clock better than a p3 even on a legacy game like quake 3.