Author: Jorge Pichard
Date: 06:17:45 06/02/04
Go up one level in this thread
On June 02, 2004 at 09:13:03, Uri Blass wrote: >On June 02, 2004 at 09:00:53, Ricardo Gibert wrote: > >>On June 02, 2004 at 08:40:55, Uri Blass wrote: >> >>>I plan to buy a cheap computer with 2 processors dual PIII1000 Mhz for testing. >>>I will probably also buy more expensive computer for WCCC with one processor >>>when I may use the cheap computer for testing with ponder on. >>> >>> >>>I still did not buy it and before buying it I have a question. >>> >>>The question is if I can tell the program which processor to use or tell the >>>program to choose automatically one of them when I have no control about it. >>> >>>Another question is if I can be sure that the processors are of equal speed(if >>>they are not of equal speed then the question which processor the program is >>>using may be important. >>> >>>Can I be sure that one 1 Ghz is not 1.01ghz when the second 1ghz is 0.99 ghz? >>> >>>Uri >> >>This would make make one processor a little over 1.02 times faster for a less 3 >>elo advantage if it assumed that each doubling of speed is worth 100 elo. To >>resolve such a difference in your testing would require an obscene number of >>games. In other words, it's not relevant. > >I agree that it is not a big difference but if there is a simple way to avoid >noise it is better to do it and I prefer to be able to control the processor >that the program is using so I do not care about difference in speed when I >always use processor A for movei and processor B for the opponents. > >Uri Since the difference in speed if there happens to be a significant difference is so minimal, simply switch from processor A for Movei to processor B to the Opponent and vise versa after half way of your testing. Jorge
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.