Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: MMX assembler for bitboards

Author: Dezhi Zhao

Date: 08:44:56 11/08/99

Go up one level in this thread


On November 08, 1999 at 11:13:01, Gunnar Andersson wrote:

>On November 08, 1999 at 10:47:51, Brian Richardson wrote:
>
>>I would suggest simply waiting for IA-64 (Merced) systems with 128 64bit
>>registers.  Why fool around with MMX ?  With that many "regular" registers (plus
>>even more "other" registers), I would rather have the compiler optimize things
>>than sweat assembler details, never mind debugging...unless you just really
>>enjoy bit twiddling.
>>
>
>The Merced is ideal, that's for sure, but I think it'll take some time before
>Merced-based compilers dominate the PC scene.
>
>Another issue is that it might take some time until the price of the Merced
>falls to a level where it'll be affordable to the likes of me. I guess this
>depends on what AMD can come up with; the Sledgehammer looks interesting.
>
>
>>Check out Intel's web info on IA-64 and download the instruction set, which
>>looks pretty good, other than missing bsr/bsf
>
>Those instructions will be sorely missed by many chess programmers...
>
>/ Gunnar

Yes for sure. But I wonder how Merced handle those instructions for so
many existing x86 programs?

dzhao



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.