Author: Peter Collins
Date: 16:42:30 10/08/03
Go up one level in this thread
On October 08, 2003 at 19:21:09, Christophe Theron wrote: >There have been several hundreds of analysis posted by many different engines >lately. > >That is very nice and we love to see all these results. > >But the ONE thing that make them hard to compare is that 90% of the posters love >to mention their name, the *town* they are living in, the amount of hash table >they are using BUT DO NOT EVEN CARE TO MENTION THE PROCESSOR THEY ARE USING and >its clock frequency in MHz!!! > >There are huge differences among the machines people are using, and I have seen >results computed on K6-2 450 and some other computed on Athlon 3000+ (one post >recently even referred to an analysis on Pentium 120MHz). > >Please remember that you are probably not all using the same kind of computer >and that there are huge differences in computing speed between these computers. > >PLEASE do not forget to mention the processor and clock speed in your analysis. > >If you don't, your results are losing a lot of their value. > >Thank you for your attention on this matter! > > > > Christophe This post reminds me of the dilemma I had 3 weeks ago, should I buy two cheaper XP 2200+ systems with each having 256MB or one 2800+ system with more ram (1.5GB) for about the same money? My answer would bore most of you, but I was delighted to see that this one faster system was more useful in that Silman position I had posted by a factor that exceeded any estimation based on Fritzmark. (I am just using it as a rough guide...) Viz., in some positions, the faster system came up with moves that suggested that it would take 10-15 times the time using the slower system... I'm happy to have splurged on that extra ram. In a nutshell, the relationships are not all linear. (Sorry for my bad grammar.)
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.