Author: Uri Blass
Date: 23:38:23 01/28/01
Go up one level in this thread
On January 29, 2001 at 02:25:46, Severi Salminen wrote: >>We can probably say that Crafty will do better against nimzo in 1 hour per game >>on athlon1.2 and not in 2 hours per game on athlon600 so we can say that Crafty >>likes more the faster hardware when we do the right comparison. > >But remember this: there are other factors than NPS. One is branching factor. >Let's assume we habe programs A and B. A has average BF of 2 and B has 6. It is >very likely that the NPS rate of A is much smaller than B's because doing more >selectivity seems to "slow" things down in terms of NPS. But program A needs >only 2x speed or time to reach an additional ply whereas B needs 6x the time or >speed. For two additional plies the figures are 4x and 36x. Even if doubling >speed results 3x increase in NPS for B and only 2x for A we can see that A earns >more from speedup in terms of plies - in terms of strength, who knows. So again >NPS doesn't tell you much. > >Severi If you do the right comparison then it is clear that B can get better results with faster hardware. You only need to do the right comparison and use faster time control on the faster hardware so A will calculate the same number of nps per move when B will calculate more nps per move. If we agree that for all programs more nps per move is better than it is clear that B can get better result on faster hardware when you use the right time control that is different for both hardwares. Uri
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.