Author: Steve Timson
Date: 14:47:41 03/11/99
I ran crafty's "bench" on my new PIII 500 and many of the various computers at my office. The results are very interesting (see below). The amazing thing is the Celeron 400. For this test it performs nearly as well as the pIII 500 and outperforms everything else including the pII 450. Also notice the serious speed decrease for win98. The interesting thing is that "98 lite" is able to reclaim the speed difference (http://www.98lite.net/). 98 lite removes some of win98's slower components and uses win95's shell/etc. Note, I do not endorse this technique. I have no idea about what millions of compatability problems this may introduce. My coworker who uses it loves it though. If you try it and it nukes your computer, don't blame me! I personally have not and will not be installing it on any of my systems. This is a purely virgin crafty 16.5 (wcrafty-16_5.exe) - no custom settings/etc. All of the computers except the celeron are Dells. The celeron is from a local vendor. Obviously the computers have different RAM, different motherboards/etc - so the take the results as you will. I still think they are pretty interesting. - Steve ------------------------------------------------------------------------------ PIII 500Mhz - 256MB - Win98 Total nodes: 59851701 Raw nodes per second: 256874 SMP time-to-ply measurement: 3.376812 Celeron 400Mhz - 64MB - Win 98 Total nodes: 59851701 Raw nodes per second: 243299 SMP time-to-ply measurement: 3.565217 PII 450 Mhz - 256MB - Win 98 Total nodes: 59851701 Raw nodes per second: 212240 SMP time-to-ply measurement: 4.086957 PII 400 Mhz - 256Mhz - Win 98 Total nodes: 59851701 Raw nodes per second: 189404 SMP time-to-ply measurement: 4.579710 PII 266 Mhz - 128 MB - "98 Lite" Total nodes: 59851701 Raw nodes per second: 136336 SMP time-to-ply measurement: 6.362319 PII 266 Mhz - 128 MB - Win 95 Total nodes: 59851701 Raw nodes per second: 132415 SMP time-to-ply measurement: 6.550725 PII 266 Mhz - 256 MB - Win 98 Total nodes: 59851701 Raw nodes per second: 118518 SMP time-to-ply measurement: 7.318841
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.