a test of how well cm11 scales from one processor to two

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

ozziejoe
Posts: 811
Joined: Wed Mar 08, 2006 10:07 pm

a test of how well cm11 scales from one processor to two

Post by ozziejoe »

I took a look at how well the king scales from one processor to two. It looks like it gets about 60 points, which I think is quite good scaling.

(Note: ratings of non-king engines not to be trusted since they only played the king and not each other; Note 2 ; I only used single processor version of rybka because the mp version would be much too strong)




Program Elo + - Games Score Av.Op. Draws

1 Rybkav2.3.2a.w32 : 3018 66 62 100 82.5 % 2749 29.0 %
2 Fruit-051103 : 2864 61 60 100 66.0 % 2749 28.0 %
3 EngineShredder10UCI : 2841 58 57 100 63.0 % 2749 32.0 %
4 Spike1.2 : 2830 55 55 100 61.5 % 2749 37.0 %
5 Naum : 2801 55 55 100 57.5 % 2749 37.0 %
6 King350_dual : 2781 33 33 300 40.8 % 2845 32.3 %
7 Ruffian_210 : 2735 58 58 100 48.0 % 2749 28.0 %
8 king350_single : 2722 34 34 300 33.0 % 2845 31.3 %


Details of tourni:

Level: Blitz 1/1
Hardware: Dual Intel(R) Core(TM)2 CPU T7400 @ 2.16GHz with 2,048 MB Memory
Operating system: Microsoft Windows XP Professional Service Pack 2 (Build 2600)

Noomen/chessbase 25 game opening testset.
Chris Tatham
Posts: 142
Joined: Wed Mar 08, 2006 8:15 pm

Re: a test of how well cm11 scales from one processor to two

Post by Chris Tatham »

Hi Joseph

Thanks for this interesting test - it is reassuring that the improvement seems reasonably consistent between different time controls and test conditions. Which version of Naum did you use?

Chris
ozziejoe
Posts: 811
Joined: Wed Mar 08, 2006 10:07 pm

Re: a test of how well cm11 scales from one processor to two

Post by ozziejoe »

it was naum 2.0


best
J