Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Wouldn't the SSDF get more accurate results if a "DEEP" program .....

Author: Joshua Lee

Date: 19:14:48 07/20/01

Go up one level in this thread


On July 20, 2001 at 18:07:21, Uri Blass wrote:

>On July 20, 2001 at 15:51:47, Joshua Lee wrote:
>
>>Wouldn't the SSDF get more accurate results if a "DEEP" program was on a dual
>>700Mhz machine than on a single 1200Mhz one?
>
>Most buyers do not use more than one processor so the interesting question is
>the rating of programs on one processor.
>
>700*2 is not more or less accurate than 1200 but simply different.
>

How different, In what ways are Deep Fritz on a dual to Deep Fritz on a Single
Processor different if both computers are running at roughly the same speed?



>>The logic is as follows the programs were intended for 2 or more cpu's
>>if you get 1.7x when you use 2 cpu's as to one
>> 700 x 1.7 = 1190 this way you still get roughly equal cpu power and the "Deep"
>>program get's to run as it was intended.
>
>Deep Fritz is intended also for one processor and if you talk about Deep Junior
>then the ssdf does not use it in their tests.
>
>Uri

This is pretty strange shouldn't they test Shredder and Junior on as well with
each version Deep and Regular?

Not only that shouldn't they test Wchess,Zarkov,Mephisto,Chess Genius it makes
sense as these programs did well against Deep Thoughtthe point being if
Program Deep Fritz Beats Wchess,Zarkov,Mephisto,Chess Genius Deep Fritz is as
Strong if not better than Deep Thought?  Isn't this the point of a ratings list
to begin with?

Then again Program X can be beaten by Wchess,Zarkov,Mephisto,Chess Genius but
beats Deep Fritz    gee this isn't an exact science.

Your thoughts...



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.