Author: Vincent Diepeveen
Date: 03:21:44 12/07/99
Go up one level in this thread
On December 06, 1999 at 16:18:39, Dann Corbit wrote: >On December 06, 1999 at 08:39:37, Vincent Diepeveen wrote: >[snip] >>More registers kick butt bigtime (unlikely that >>intel changes this till the merced comes out as this would require new >>compiler and incompatibility with current software), but i have no >>idea how much it would speedup my program. >Except when you change context from thread to thread. That why the 256 register >SPARC is handicapped for multiple CPU's. A context switch is very expensive. >Imagine a million registers and going from thread to thread for an easy mind >experiment about why lots of registers is not always a good thing. >[snip] pushing a few registers won't be a problem. i didn't test at multiprocessor sparcs, i would have like that though... ...alpha seems also a lot slower at more than 1 cpu btw. single cpu 667Mhz 21264 experimental at icc: 850k nps dual cpu 667Mhz 21264 experimental: 1M to 1.2M nodes a second (crafty of course)
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.