SI
SI
discoversearch

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)
AMD 210.78-4.8%Dec 12 9:30 AM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Daniel Schuh who wrote (10019)9/23/2000 2:24:19 PM
From: CirruslvrRead Replies (3) of 275872
 
Daniel and Andreas - RE: Swim benchmark in SPECfp2000

Here is a comparison of the different benchmarks within SPECfp2000 and the % difference between the scores of a 1GHz PIII on the 820 and 840 chipset. The 840 has double the memory bandwidth of the 820, equal to Willy's. The numbers are from the base ratio. I didn't know whether to use base ratio or ratio so I just chose base ratio. The difference between the two is not a whole lot...


VC820 OR840 % diff
168.wupwise 400 422 6%
171.swim 408 493 21%
172.mgrid 248 279 13%
173.applu 248 278 12%
177.mesa 508 512 1%
178.galgel 292 311 7%
179.art 376 410 9%
183.equake 232 256 10%
187.facerec 300 313 4%
188.ammp 285 298 5%
189.lucas 304 329 8%
191.fma3d 278 296 6%
200.sixtrack 175 176 1%
301.apsi 343 366 7%


As you can see, the swim benchmark, which C't said Willy "explodes" on, gains the most from the extra memory bandwidth.

When Willy comes out Intel may want to show comparisons of the PIII on the 820 chipset (1.6 GB/s memory bandwidth) and Willy on the 850 chipset (3.2GB/s) . Good thing we have 840 (3.2GB/s) benchmarks to do some REAL comparisons with! Sixtrack would be a good benchmark to compare the FPUs of the two processors since the gain from 820 and 840 is fairly negligible.

spec.org
spec.org
spec.org
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext