Author: Dann Corbit
Date: 19:39:29 02/18/03
Go up one level in this thread
On February 18, 2003 at 20:58:09, Sean Mintz wrote: >Did a new compile with better options for intel. > >New results: > >Compiler MIPS >Intel C 5 [1] 1452.4 >Intel C 5 [2] 1470.1 >Intel C 7 [1] 1354.2 >Intel C 7 [2] 1331.6 >GCC [3] 1598.1 > >1 - (Profile directed) -O2 -Qipo -QxiM -G6 -Qunroll -Qprof_use > >2 - (Profile directed) -O3 -G6 -QxiMK -Qunroll -Qipo -Ob2 -Qprof_use > >3 - -O3 -march=athlon-mp -mcpu=athlon-mp -mfpmath=sse -mmmx -fomit-frame-pointer >-funroll-loops -finline-functions -fprefetch-loop-arrays For Intel, I usually use a set of flags similar to this: /nologo /G6 /ML /W5 /Ob2 /FD /G7 /GA /GF /Gs /O3 /QaxiMKW /Qip /Qwp_ipo /Zm500 /Qprof_genx /c Followed by: /nologo /G6 /ML /W5 /Ob2 /FD /G7 /GA /GF /Gs /O3 /QaxiMKW /Qip /Qwp_ipo /Zm500 /Qprof_use /c after profiling. Exactly what version of Mingw's GCC are you using? The performance I get usually stinks in comparison to what you are getting.
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.