Author: Robert Hyatt
Date: 05:09:49 01/28/00
Go up one level in this thread
On January 28, 2000 at 03:48:54, Tom Kerrigan wrote: >On January 28, 2000 at 01:00:23, Robert Hyatt wrote: > >>>I remember that MChess was running at ~2,500 NPS on a 133MHz Pentium in 1995. >>>133M / 40k = 3,300 >>3-4K seems to be a good number. Frans has this down to 2K. A program like >>Hiarcs seems to be at 10K or so, although it is impossible to say what he >>is doing, lots of eval or lots of computation trying to figure out which >>parts of the tree to extend further... > >I know a few people who have seen MChess's evaluation function. They say it's >miles long and evaluates everything under the sun. That's why I suspect the >pruning is mainly based on the eval. function, but that's just a guess. > >>OK.. assume 40K is _the_ answer for an X86 program to search one node. What >>do we know? Eval instructions? search instructions? IE we _still_ don't >>know what that means. > >From the abstract: "On a general-purpose computer, the computation performed by >the chess chip for one chess position is estimated to require up to 40,000 >general-purpose instructions. At 2 to 2.5 million chess positions per second, >one chess chip is equivalent to a 100 billion instructions/sec supercomputer." > >So it doesn't explicitly say, but the final conclusion suggests that it's 40k >instructions per node (including search, eval, etc.). > >Even if the number doesn't include search, I doubt it would have much effect on >the 40k figure. I have seen people claim to search a node w/ 1000 instructions. > >Also from the abstract: "Each of these chess chips contains one of the most >sophisticated chess evaluation functions ever designed, whether in hardware or >in software." > >So Hsu never says it's THE most sophisticated. If you think about it, saying >your function is THE most sophisticated is not smart. How do you know there >isn't some unknown guy in, say, Bangladesh, with a bigger function? > >-Tom As I said... he _believes_ it is the best, because of the test games he played vs micros to test this feeling. But in writing, it is logical to be conservative so that you don't have to retract something once someone proves your statement was "too much"...
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.