Author: Robert Hyatt
Date: 11:27:26 11/19/02
Go up one level in this thread
On November 19, 2002 at 14:17:42, Gian-Carlo Pascutto wrote: >On November 19, 2002 at 14:11:18, Robert Hyatt wrote: > >>On November 19, 2002 at 12:25:11, Gian-Carlo Pascutto wrote: >> >>>On November 19, 2002 at 11:35:24, Robert Hyatt wrote: >>> >>>>Bitboards have a bit of a performance advantage on 64 bit processors, >>> >>>Proof? >>> >>>-- >>>GCP >> >>Counter-proof? >> >>Seems intuitively obvious to me. > >It seems intuitively obvious to Vincent that bitboards suck, but >I believe neither of you. Doesn't matter to me. I use them. I am happy with them. Frans apparently liked them enough to convert to them. Several others are using them with no problems. I personally don't believe _anything_ Vincent says, but that is a different issue. > >>Bitmaps seem to suffer _no_ performance penalty on >>X86 with 32 bits, compared to 0x88. > >How do you determine this? By comparing the speed I generate moves at compared to 0x88, for example? All the "good generators" are within a factor of two or so of each other, depending on what they do in Make/Unmake... > >>Seems intuitively obvious to me that they will pick >>up speed on a machine that does 64 bit operations. > >Not so obvious to me, I've posted a few issues with it >here before. > >>Bruce and I did this comparison when he used the alpha in the WMCCC (1997 I >>think). >> >>He re-compiled ferret for the alpha, did the same for Crafty. My speed >>improvement was >>significantly better than his on the _same_ machine. Because of the 64 bit >>stuff. His program didn't need any 64 bit stuff so it was wasted... > >Doesn't say a thing. The programs are much more different than just >bitboard vs nonbitboard. That's simple hand-waving. It shows that a 64 bit program gains more from going to a 64 bit architecture than a 32 bit program does. I'm not about to make this into a computer architecture 101 discussion. There are plenty of good books around that explain _exactly_ why this is true. I don't see any nead to repeat those arguments over and over... If you don't want to believe them, feel free. It must mean _everybody_ in the computer architecture field is wrong, however... > >-- >GCP
This page took 0.01 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.