Author: Gerd Isenberg
Date: 23:50:23 08/26/03
Go up one level in this thread
On August 26, 2003 at 20:11:46, Dan Andersson wrote:
> Depends on what you use it for. It's slower than SSE but last time I looked the
>precision was better. It was quite a while ago. But maybe a rewrite to SSE might
>be faster anyway.
> It is probably time to retire the archaic FP cludge x87. As well as 3Dnow! and
>MMX. Even though I really liked the 3Dnow! architecture. Far superior to the
>MMX/SEE IMO.
For FP i agree. But there are simply eight 64-bit MMX registers, there are fast
direct path instructions like movq2dq.
My intention was to use XMM and MMX for KoggeStone fill algortiths.
MMX is nice to generate propagators - without polluting XMM.
BitBoard FillUpOccluded(BitBoard g, BitBoard p) // generator, propagator
{
g |= p & (g << 8);
p &= (p << 8);
g |= p & (g << 16);
p &= (p << 16);
return g |= p & (g << 32);
}
Ok, the only usefull thing for MMX is to hold some 64-bit constants ;-)
Regards,
Gerd
>
>MvH Dan Andersson
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.