Author: Gerd Isenberg
Date: 02:37:18 12/05/03
Go up one level in this thread
On December 05, 2003 at 00:22:11, Russell Reagan wrote: >I found an old piece of code I had written to test the speed of a few >bitscanning routines (ex. FirstOne() in Crafty). I was curious how different >they would run on a 64-bit machine, so I sent them to Dr. Hyatt and he was kind >enough to run them for me on the Opteron. Thanks Bob ;-) <snip> > >Here are some short descriptions of the different bitscans: > >magic bitscan >Created by (I think) Matt Taylor, or maybe it was Gerd Isenburg. It uses de >Bruijn sequences (in the form of a "magic" number). Really cool. Hi Russell, No, seems that Matt Taylor did a kind of reinvention: Using de Bruijn Sequences to Index a 1 in a Computer Word Charles E.Leiserso Harald Prokop Keith H, Randall MIT Laboratory for Computer Science, Cambridge, MA 02139, USA July 7, 1998 > >table 16 >I got this one from Dann Corbit, who (I think) got it from Beowulf. It uses a >16-bit lookup table, which kind of makes me cringe on a machine with a small >amount of cache, but the Opteron has 1 MB of L2 so maybe we can live with it. > >gerd >I got this one from (I think) Gerd Isenburg, or maybe it was Matt Taylor. It >uses the same magic number concept, but in a way that is intended to be more >32-bit friendly, so it is not suprising that it is slower than the magic bitscan >on 64-bit hardware. Honor to honor is due - definitely Matt Taylor. He invents the magic folding trick for the 32-bit de Bruijn multiplication! So please call the routine matt ;-) Cheers, Gerd
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.