Author: Bruce Moreland
Date: 19:13:23 01/18/00
Go up one level in this thread
On January 18, 2000 at 04:52:53, John Warfield wrote: >2r2rk1/p3bb1p/2n1Q1p1/q2pP3/3P1P2/p1NB1NR1/1P4P1/1K1R4 w - - id Evans.L - >Pilnick,C; bm d3g6; > > > In the last printed computer chess reports magazine this position was used as >a computer bench mark. Various programs were listed along with how fast they >solved the above problem. I no longer have the article, but if memory serves >Cray blitz solved this in 1 second. Fritz4 was the fastest to solve it out of >the Pc programs I think on a pent 133 it solved it in 2 min which was by far the >fastest out of the pc program. On my K6-350 fritz5.32 solves it in 32sec!! Using >some of my other programs the position wasn't solved. Looks like fritz is the >best at tactics, chessmaster 6000 took a little over 2 min to solve. According >to the original article some of the older programs like cm3000 could not solve >the problem even after thinking for days on the latest hardware at the time, >which proved that programs had indeed advanced in sophistication as well as >hardware speed. Does anyone have the old article? I was interested in the times >for fritz and genius. To compare just how much programs have advanced since >then. On a P6/200 (a slow machine, relatively) mine takes 3 minutes to find this, resolving the fail high after a little under four minutes at -1.78. First positive score is +1.31, in the next ply, returned after 7 minutes. First crush is +4.48 after an hour and a half. bruce
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.