Author: Ernst A. Heinz
Date: 08:05:24 07/19/98
Go up one level in this thread
On July 17, 1998 at 18:28:53, Christophe Theron wrote:
>On July 17, 1998 at 17:11:01, Ernst A. Heinz wrote:
>
>>On July 17, 1998 at 14:44:23, Christophe Theron wrote:
>>>
>>>If we agree on "400MHz Alpha roughly as fast as 300MHz Intel for chess
>>>programming", we are still comparing the programmer's skills.
>>
>>If you want to define something as cloudy as "nearly equal performance
>>for chess programming" across different platforms, my experience
>>suggests that the well-known and independently established SPECint95/98
>>ratings serve this purpose better than anything else.
>>
>>=Ernst=
>
>Ernst, would you support the idea of "nearly equal platform", at least in
>principle?
Yes, of course -- but it strongly depends on the event and how it is called.
I do not like the idea of limiting the official WCCC/WMCC events any further.
IMHO, their rules and restrictions on hardware are fine.
However, if you come up with an event that aims to compare mainly the software,
(e.g. "World Chess Program Championship") it makes perfect sense to limit the
computing power (e.g. SPECint95/98) of the participating hardware to some
*fast* state-of-the-art range (i.e. the limit should not be set too low).
>Do you have SPECint results for Alpha and Pentium? Do you know where I can get
>those results? Thanks.
Alpha-21164a 500MHz ~15 SPECint95
600MHz ~18 SPECint95
AMD K6-1 266MHz ~11 SPECint95
HP PA-8200 220MHz ~16 SPECint95
Intel PPro 200MHz ~ 9 SPECint95
Intel P-II 333MHz ~13 SPECint95
350MHz ~14 SPECint95
400MHz ~16 SPECint95
MIPS R10000 250MHz ~15 SPECint95
Sun UltraIIi 333MHz ~14 SPECint95
Right now, I deem ~15 SPECint95 as a reasonable choice for something like a
"World Chess Program Championship" (i.e. allow 500MHz Alpha and <400MHz x86).
=Ernst=
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.