Author: Robert Hyatt
Date: 20:41:08 03/14/03
Go up one level in this thread
On March 14, 2003 at 18:54:24, Jeremiah Penery wrote: >On March 14, 2003 at 12:14:32, Robert Hyatt wrote: > >>I am attaching the SMP results for 1, 2 and 4 cpus, for the first 12 BS2830 >>positions. >>Vincent apparently wanted to pick and choose to produce a set of positions >>Crafty does >>badly on. I prefer to take a random set of positions, in this case the first 12 >>BS2830's >>without any attempt to "pick the best or the worst." > >>Speedup per position: >> >> 2cpu 4cpu >> avg 1.8 2.7 >> >>The above average weighs every position equally. An alternative >>way is to add all the times in the columns, and then divide the >>one cpu total by the two cpu total (or the 4 cpu total) which >>weights the positions that took longer, higher. That gives the >>following speedup numbers. >> >> 2cpu 4cpu >> avg 1.6 2.8 >> >>If I were trying to do this for a paper, I would have run the >>test a bit differently, because at least 2 of the 12 positions >>were not searched deeply enough for reasonable parallel comparison, >>since 6 and 7 took .4 seconds each, compared to almost three minutes >>for number 2. However, I didn't think Vincent was worth going to >>a lot of trouble, since he has these sorts of fantasies from time >>to time. > >Your data doesn't address Vincent's claims at all, unless you just failed to >mention it. Vincent wanted comparison of SMP speedup with asymmetric king >safety on and off. You made no mention of asymmetry in your post, so there's no >way to tell what your data is supposed to show. Reread what he wrote. He specifically said do _not_ run a non-asymmetric version of Crafty. But, in fact, there is little to argue with him about, since he only rants and raves without offering a single thing to discuss reasonably. He claimed the speedup was 1.4 or worse. In fact, it was 1.8 for two, for this particular set of 12 positions. I'd be happy to run a non-asymmetric version if you think that would be interesting...
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.