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: pgerassi who wrote (101411)	4/1/2000 2:32:00 AM
From: Ali Chen	Read Replies (1) \| Respond to of 1572124

Pete, <. the page shown by AMD does not seem impossible..> Do not waste your time arguing with Elmer FUDD. He is here for reasons other than taking and making logical arguments. Of course, it is absolutely ridiculous to accuse a $3B company with 13,000+ employees in falsifying benchmark scores. More, Elmer was pointed out on several occasions that the German c't magazine also has fairly low scores for Cumine when compiling SPEC benchmarks with really available IFL2.4 compiler and affordable hardware. Of course, it may be debated whether it is valid to compare two processors in identical environment as AMD did, or compare current but unavailable and non-affordable best system against AMD system built from cheap commodities. (Note, that is the reason he is so "sensitive" to Dell notes about long lead times on 800MHz system, and 1000Mhz is nowhere in the sight) BTW, the more he post about benchmarks, especially for FP, the more stupid he will look one day...

To: pgerassi who wrote (101411)	4/1/2000 10:41:00 AM
From: Elmer	Read Replies (1) \| Respond to of 1572124

Re: "Dear Elmer: None of those were on a BX chipset using 4.0 with 256MB PC100. They were also using what appears to be a 840 with Dual RMBS PC800 DIMMS. SPECbaseFP uses SPEC Specified switches. They seem to have been allowed more latitude than what I thought should be allowed given what I read at their site. Peak is allowed to have different switches per section. None of those runs use a non Intel FORTRAN compiler. Here is a page for a 600MHZ Pentium III BX PC100 v2.4 spec.org. If you use a 4/3 ratio to est 800 MHZ BX you get 19.8 95basefp An Athlon 600 MHZ 751 PC100 V2.4: spec.org. So the page shown by AMD does not seem impossible given the platform and compiler. All the ones you show use later compiler and Rambus with (840) (Dell, Compaq, and Siemens do not say which chipset). Pete " Pete, that was a Katmai version of PIII without ondie L2. The CuMine version, with ondie L2, shows dramatic improvement over it's older version. Therefore AMD's CuMine benchmarks remain unexplainably low and AMD is not helping out when they withold the compiler switches. One would want to think that this is just carelessness on AMD's part however I don't think Intel would be given the same latitude under similar circumstances. EP