Author: Ernst A. Heinz
Date: 05:02:43 06/16/98
Go up one level in this thread
On June 16, 1998 at 03:41:50, Stefan Meyer-Kahlen wrote: > >On June 16, 1998 at 00:59:00, Jouni Uski wrote: > >>I wonder if anybody has still achieved 1 000 000 NPS speed in PC software! >>With 400 MHZ prosessors may be somebody has seen it - even in some special >>endgame position? I am just curious... >> >>Jouni > >DarkThought did it in an endgame position on an Alpha-767 in Paris. Probably >Ernst can give us more details. > >Stefan Sure I can -- however, most people do not consider an Alpha to be a PC ... Our WWW page at <http://wwwipd.ira.uka.de/Tichy/DarkThought/> mentions that the current "DarkThought" peaks at 900K nps on a 600MHz Alpha-21164a when searching position ``8/8/7K/8/6p1/6k1/4R3/8 w - - 0 1'' from ECE3 with a fixed time of 5 sec. The very same position and settings let the WMCC'97 version of "DarkThought" break the 1M nps barrier in Paris while running on a 767MHz Alpha-21164a. Today, "DarkThought" reaches speeds of >= 800K nps in many endgame positions with little material on a 600MHz Alpha-21164a such that 1M nps should no longer be that exceptional on a 767MHz Alpha-21164a. As soon as the new Alpha-21264 CPUs arrive we hope to see 1M nps already in late middlegames. Our 1M nps achievement in Paris is documented in the ICCA Journal: J. van denHerik. Surpassing the speed limit of 1,000,000 nodes per second (photograph of DarkThought versus Toledo 2000). In ICCA Journal, Vol. 20, No. 4, page 273, December 1997. T. Hamlen and M. Feist. Report on the 15th World Microcomputer Chess Championship. In ICCA Journal, Vol. 20, No. 4, pages 254-256, December 1997. =Ernst=
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.