terminat0r
Silver Member | Редактировать | Профиль | Сообщение | Цитировать | Сообщить модератору Процессоры processor : 15 vendor_id : AuthenticAMD cpu family : 15 model : 33 model name : Dual Core AMD Opteron(tm) Processor 885 stepping : 2 cpu MHz : 2600.000 Компиляторы Intel(R) Fortran Intel(R) 64 Compiler Professional for applications running on Intel(R) 64, Version 11.1 Build 20091130 Package ID: l_cprof_p_11.1.064 gfortran: Target: x86_64-suse-linux, gcc version 4.4.1 [gcc-4_4-branch revision 150839] (SUSE Linux) результаты смотреть третью колонку (время в секундах) no patch 1) ifort -O3 index r_index time --------------------------------------------- 790 7.8900000000E+001 7.3572598000E+001 791 7.9000000000E+001 7.3612601000E+001 792 7.9100000000E+001 7.4864679000E+001 793 7.9200000000E+001 7.4824676000E+001 794 7.9300000000E+001 7.4168635000E+001 795 7.9400000000E+001 7.4224639000E+001 796 7.9500000000E+001 7.8480905000E+001 797 7.9600000000E+001 7.4980686000E+001 798 7.9700000000E+001 7.4216638000E+001 799 7.9800000000E+001 7.4856678000E+001 800 7.9900000000E+001 7.4232640000E+001 patch 2) ifort -O3 790 7.8900000000E+001 8.5125321000E+001 791 7.9000000000E+001 8.4549284000E+001 792 7.9100000000E+001 8.5925370000E+001 793 7.9200000000E+001 8.5245327000E+001 794 7.9300000000E+001 8.5181324000E+001 795 7.9400000000E+001 8.5229326000E+001 796 7.9500000000E+001 8.5341334000E+001 797 7.9600000000E+001 8.3949246000E+001 798 7.9700000000E+001 8.4545284000E+001 799 7.9800000000E+001 8.5209325000E+001 800 7.9900000000E+001 8.3813239000E+001 patch 3) ifort -O3 -ip -ipo -static -xP -finline-functions -funroll-loops -fp-model precise -ftz 790 7.8900000000E+001 5.0359147000E+001 791 7.9000000000E+001 4.9567097000E+001 792 7.9100000000E+001 4.8391025000E+001 793 7.9200000000E+001 4.8415025000E+001 794 7.9300000000E+001 4.8359022000E+001 795 7.9400000000E+001 4.7958998000E+001 796 7.9500000000E+001 4.8703044000E+001 797 7.9600000000E+001 4.7942996000E+001 798 7.9700000000E+001 4.8783048000E+001 799 7.9800000000E+001 4.8395025000E+001 800 7.9900000000E+001 4.8363023000E+001 no patch 4) ifort -O3 -ip -ipo -static -xP -finline-functions -funroll-loops -fp-model precise -ftz seg fault no patch 5) ifort -O3 -xSSE3 seg fault patch 6) ifort -O3 -xSSE3 works! 7) gfortran -O3 -march=amdfam10 -funroll-all-loops -fprefetch-loop-arrays 790 7.8900000000E+001 6.3783986000E+001 791 7.9000000000E+001 6.2195887000E+001 792 7.9100000000E+001 6.3231952000E+001 793 7.9200000000E+001 4.8939058000E+001 794 7.9300000000E+001 4.5678856000E+001 795 7.9400000000E+001 4.4110756000E+001 796 7.9500000000E+001 4.4118757000E+001 797 7.9600000000E+001 4.6326895000E+001 798 7.9700000000E+001 4.8199013000E+001 799 7.9800000000E+001 4.9611100000E+001 800 7.9900000000E+001 4.3370711000E+001 | Всего записей: 2084 | Зарегистр. 31-03-2002 | Отправлено: 12:13 22-04-2010 | Исправлено: terminat0r, 12:53 22-04-2010 |
|