Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Pentium IV mobile processors have very good results ...

Author: Frank Quisinsky

Date: 05:34:53 01/24/04

Go up one level in this thread


On January 24, 2004 at 07:30:28, Aaron Gordon wrote:

>On January 24, 2004 at 05:41:15, Frank Quisinsky wrote:
>
>>On January 23, 2004 at 18:03:48, enrico carrisco wrote:
>>
>>>On January 23, 2004 at 17:25:21, Frank Quisinsky wrote:
>>>
>>>>On January 23, 2004 at 12:02:42, Thomas Mayer wrote:
>>>>
>>>>Hi Thomas,
>>>>
>>>>>you write that nonsense all the time -> maybe you READ your forum ?
>>>>
>>>>http://f22.parsimony.net/forum41668/messages/24801.htm
>>>>
>>>>Best
>>>>Frank
>>>>
>>>
>>>Message from link above:
>>>
>>>"Hi Frank,
>>>
>>>Axon Bench has Geniss Axon XP chess engine, totaly written in machine language.
>>>It is mixture of 16-bit and 32-bit instructions. So, we could say that Athlon
>>>processors are more optimized for that kind of code compared with Intel
>>>machines. The test proves that for computer chess AMD machines are the best
>>>choise. PIV-s are simply over-optymized only for CLEAR 32-bit code. It is
>>>interesting fact that huge number of chess engines (with exception of Fritz)
>>>runs much better on AMD processors !
>>>
>>>regards, Vladan"
>>>
>>>This is a little different than saying it is optimized for AMD.  It's a fine
>>>line perhaps.  Saying Axon is optimized for AMD would be like saying all chess
>>>engines are optimized for AMD because they simply run faster on those CPUs.
>>>
>>>-elc.
>>
>>Hi,
>>
>>this is right but ...
>>
>>The Axon Bench test is after my information a part from Axon chess engine. Axon
>>chess engine is working on Athlon processors faster! Gandalf 4.32x versions or
>>Crafty works after my information not faster on Athlon processors. A good
>>example is Anaconda too. Anaconda is working on Athlon processors faster.
>>
>>And this is the reason for me to write (Axon as a part of a chess program which
>>is faster on Athlon processors) that Athlon processors will have a better result
>>as Intel processors with Axon bench test. We can start the Crafty bench and
>>compare the results on Intel or Athlon and the results are completly different.
>>
>>Not more not less ...
>>
>>Best
>>Frank
>
>Crafty still rules on Athlons too.
>Here are some 18.11 results I have from back in the day:
>
>CPU type           Nodes/sec
>
>AthlonXP 2.52GHz   1,578,197
>AthlonXP 2.2GHz    1,370,541
>Pentium 4 3.25GHz  1,335,397
>AthlonXP 1.86GHz   1,197,253
>Tbird 1.68GHz      1,041,610
>Pentium 4 2.47GHz    992,009
>Athlon Tbird 1.3GHz  760,299
>Pentium 4 1.7GHz     687,406
>Celeron (P3) 1GHz    575,475
>Pentium 4 1.5GHz     569,185
>
>Even the old P3 cored Celeron 1GHz is faster than a P4-1.5GHz. IMO P4 is a waste
>of money. Runs hotter, is slower, and to get a CPU that comes close to my $50
>chip (1.73 @ 2.5ghz) you'd need to pay $1000. THEN you'd have to overclock that
>$1000 chip and it would now dump close to 200 watts. Then you get to pay $200+
>for a cooling system for it. ZZZzzz....

Hi Gorden,

this is interesting for me ...
In this case my information is wrong.
At the moment I don't used Athlon PCs but the results from the Athlons are a lot
stronger.

Thanks for your information!

Best
Frank




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.