由买买提看人间百态

topics

全部话题 - 话题: pgcc
(共0页)
j**u
发帖数: 6059
1
来自主题: Computation版 - pgcc vs gcc
【 以下文字转载自 Linux 讨论区 】
发信人: aaddoo (栀子花), 信区: Linux
标 题: pgcc vs gcc
发信站: BBS 未名空间站 (Fri Feb 2 15:52:36 2007)
我现在运行的计算程序,如果用pgcc编译,计算速度要比用gcc编译的快一倍。 这个差
的也太大了。怎么样让gcc编译的计算快一点?
现在正在运行一个任务,估计要两个月,要是能快一倍,一个月就能搞定。
w****g
发帖数: 206
2
做了以下修改,还是不工作,有牛人帮看看么?
CC=pgcc
CXX=pgCC
LINK=pgCC
AR="/usr/bin/ar cru"
RANLIB=/usr/bin/ranlib
# DEPENDFLAG="-MM"
DEPENDFLAG="-MM"
# REL_FLAGS="-fPIC"
REL_FLAGS=""
SO_LIBS=""
# CPPFLAGS="-Wall"
CPPFLAGS="Minform=warn"
# OPT_CPPFLAGS="-O3 -xW -DNDEBUG"
OPT_CPPFLAGS="-O3"
# DBG_CPPFLAGS="-O0 -g -DDEBUG"
DBG_CPPFLAGS="-O0"
DO_64BITS=1
ARCH_DIR="$canonical_name"
# SYS_LIBS="-i-static"
SYS_LIBS="-Bstatic"
;;
w****g
发帖数: 206
3
【 以下文字转载自 Programming 讨论区 】
发信人: wooing (woo), 信区: Programming
标 题: 求助:有用portland compiler 的么?
发信站: BBS 未名空间站 (Mon Nov 1 21:42:17 2010, 美东)
老板让我试着用portland的compiler build code,以前一直用intel的,这两个之间的
flag是怎么对应的啊?
CC=icc
CXX=icpc
LINK=icpc
AR="/usr/bin/ar cru"
RANLIB=/usr/bin/ranlib
DEPENDFLAG="-MM"
REL_FLAGS="-fPIC"
SO_LIBS=""
CPPFLAGS="-Wall"
OPT_CPPFLAGS="-O2 -DNDEBUG"
DBG_CPPFLAGS="-O0 -g -DDEBUG"
DO_64BITS=1
ARCH_DIR="$canonical_name_guess"
SYS_LIBS="-i-static"
;;
... 阅读全帖
w****g
发帖数: 206
4
来自主题: Programming版 - 求助:有用portland compiler 的么?
老板让我试着用portland的compiler build code,以前一直用intel的,这两个之间的
flag是怎么对应的啊?
CC=icc
CXX=icpc
LINK=icpc
AR="/usr/bin/ar cru"
RANLIB=/usr/bin/ranlib
DEPENDFLAG="-MM"
REL_FLAGS="-fPIC"
SO_LIBS=""
CPPFLAGS="-Wall"
OPT_CPPFLAGS="-O2 -DNDEBUG"
DBG_CPPFLAGS="-O0 -g -DDEBUG"
DO_64BITS=1
ARCH_DIR="$canonical_name_guess"
SYS_LIBS="-i-static"
;;
象以上这些,除了吧icc改成pgcc,icpc改成pgCC外,其他的怎么修改啊?
u***o
发帖数: 385
5
来自主题: Hardware版 - i3 530 oc super_pi
把i3 oc 到 190*22=4.18G,运行稳定。不知道天热起来以后,是不是需要降一点。
~/Desktop/super_pi$ ./super_pi 20
Version 2.0 of the super_pi for Linux OS
Fortran source program was translated into C program with version 19981204
of
f2c, then generated C source program was optimized manually.
pgcc 3.2-3 with compile option of "-fast -tp px -Mbuiltin -Minline=size:
1000 -Mnoframe -Mnobounds -Mcache_align -Mdalign -Mnoreentrant" was used for
the
compilation.
------ Started super_pi run : Mon Apr 19 18:53:14 PDT 2010
Start of PI c
d********l
发帖数: 741
6
来自主题: Hardware版 - 刚才跑了一把superPI
目前还没有比我的更快的: Total calculation(I/O) time= 16.932(0.836) Sec.
CPU是i******[email protected]
------ Started super_pi run : Thu Apr 10 01:24:30 PDT 2014
------ Ended super_pi run : Thu Apr 10 01:24:47 PDT 2014
Version 2.0 of the super_pi for Linux OS
Fortran source program was translated into C program with version 19981204
of f2c, then generated C source program was optimized manually.
pgcc 3.2-3 with compile option of "-fast -tp px -Mbuiltin -Minline=size:
1000 -Mnoframe -Mnobounds -Mcache_align -Mdalign -Mn... 阅读全帖
(共0页)