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: grok who wrote (61609)	6/13/1999 1:17:00 AM
From: Elmer	Read Replies (2) \| Respond to of 1583878

Re: "The Intel people will cry foul over this if it is true. Can you imagine the capital letters we'll be seeing from Paul?" All components and software must be commercially available as a condition of publishing the SPEC results. Also note that the Xeon benchmarks are all SPECint_base95 results, or at least the SPECint_base95 results are the same as the SPECint95 results. The significance is that SPECint_base95 results must use the same compiler options for all the tests run. Other companies use different compiler options and get up to 30% better results for SPECint95 compared to SPECint_base95. AMD boasted of superior SPECint95 results. One might wonder how they would both compare in SPECint_base95. EP

To: grok who wrote (61609)	6/13/1999 1:42:00 AM
From: Ali Chen	Read Replies (2) \| Respond to of 1583878

KZ, <are the binaries produced by Intel's compiler available for AMD to use on K7?> If you would be patient to look down the bottom of your link spec.org you would find the following text: "The Intel Compiler 4.0 for Windows NT is shipped with VTune 4.0, Microsoft Visual C++ 6.0 is also installed for libraries" This means: if you buy the open-market product "VTune" (from Intel), you are set to make binaries exactly as used by Intel (since compiler optimization flags are published as well by SPEC requirements). Anyone can go and buy the "VTune" product, therefore the "binaries" can be freely reproduced for fair benchnarking. However, since there is no optimization flags yet for K7 in that Intel FORTTRAN compiler :), the only fair way for today would be to compare so-called "base" numbers, where no special treatment/optimization for sub-benchmarks is done, only few common flags for all ten FP executables.