Computer Chess Club Archives




Subject: Re: SSE2 bit[64] * byte[64] dot product

Author: Fabien Letouzey

Date: 08:35:10 07/19/04

Go up one level in this thread

On July 19, 2004 at 10:59:05, Anthony Cozzie wrote:

>32 cycles is _really_ good.  I think that on average rotated bitboard attack
>generation is 20 cycles, so that is 50 cycles / piece / mobility = 500 cycles
>(~250 ns on my computer) for all pieces, which is really not bad.  In fact, 32
>cycles is not that much slower than popcount!


Why would it be?  Maybe I misunderstand the calculation but it seems to me that
a table-based solution, similar to one for popcount, would also do the work?!

I am not saying that it would be better than Gerd's code, but my point is that I
would expect that there is a solution mearly as fast as a general popcount (not
one specific for sparse boards).


This page took 0 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.