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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Tenchusatsu who wrote (142610)	9/1/2001 1:43:58 PM
From: Windsock	Respond to of 186894

Ten - Re:"Don't pretend that SPEC95 isn't legitimate, because that was exactly the benchmark used by Dirk Meyer in his 6/10/1999 dinner presentation." The only legitimate benchmark for Dan3 is one that makes Athlon look good. Hence, the reliance on ancient measurements like SuperPi. Everybody knows how people calculate Pi in their daily work. Next thing you know, Dan3 will be touting the speed in getting a DOS prompt on DOS 3.2

To: Tenchusatsu who wrote (142610)	9/1/2001 8:12:35 PM
From: FJB	Read Replies (3) \| Respond to of 186894

I notice that the highest scoring Athlon system submitted to SPEC by AMD itself uses the Intel compiler. Apparently, AMD gets the best results using it too. specbench.org I don't understand Dan's commentary about "Real world code creation uses compilers that don't fail to compile on many applications, and are only reliable on the one targeted application". SPEC2000 runs 25 applications all together and most of them seem to by written by academics which wouldn't have a vested interest in tuning the code for any particular microprocessor. Of course Intel tunes their compiler to make code run as fast as possible on whatever the latest generation of CPU is. Isn't that what you want in a compiler? If they put in a flag for, "If Athlon detected, produce slow object code", I could see a reason for argument. But they haven't made that flag a feature of the compiler yet. :-) Bob

To: Tenchusatsu who wrote (142610)	9/1/2001 11:27:54 PM
From: Dan3	Read Replies (2) \| Respond to of 186894

Re: Don't pretend that SPEC95 isn't legitimate SPEC95 binaries compiled before P4 or Athlon were released are very legitimate for testing the processors. Have you ever heard of a double-blind test? What Intel has been doing with benchmarks is like the cigarette companies gathering together a bunch of 80 year olds who smoked all their lives and announcing "if smoking isn't safe, how come all these smokers lived to be 80?" That's why I don't trust benchmarks tuned to a processor and developed with it in hand. Binaries using algorithms that were selected after Intel spent millions of dollars and tens of thousands of man-hours figuring out what made P4 look good, (as opposed to what reflected typical use and practice) and then tweaked at an assembler level to favor P4 even further, are not valid.