GUPS Benchmark Build

Random memory performance often maps directly to application performance and application development time. A small percentage of random memory accesses (cache misses) in an application can significantly affect the overall performance of that application

http://www.sandia.gov/~sjplimp/algorithms.html에서 소스 코드를 다운받을 수 있다.

다운받은 소스 코드를 압축 해제한다.

$tar -xzvf ./gups.tar.gz

mpi library가 필요하므로, 이를 설치한다.

$sudo apt-get install libmpich-dev

Makefile.linux를 빌드 환경에 맞추어 수정한다.

#before modification
CC = gcc
CCFLAGS = -O -g -I/home/sjplimp/tools/mpich/include -DCHECK
LINK = gcc
LINKFLAGS = -O -g -L/home/sjplimp/tools/mpich/lib
LIB = -lmpich

#after modification
CC = gcc
CCFLAGS = -O -g -I/usr/include/mpich/ -DCHECK
LINK = gcc
LINKFLAGS = -O -g
LIB = -lmpich

빌드한다.

$ make -f Makefile.linux gups_vanilla

References :
[1] Giga-updates per second, Wikipedia, http://en.wikipedia.org/wiki/Giga-updates_per_second
[2] HPCC Challenge, http://icl.cs.utk.edu/hpcc/index.html
[3] Steve Plimpton’s Hompage, http://www.sandia.gov/~sjplimp/algorithms.html
[4] Steve Plimpton’s Hompage, http://www.sandia.gov/~sjplimp/download.html

+덧(2014/11/07)
그런데 왜, HPCC challenge 홈페이지에서 소스 코드를 다운받지 않고 Steve의 홈페이지에서 받는지? 둘은 어떤 관계인가?

Advertisements
Tagged with: , ,
Posted in 2) Computer Engineering

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

누적 방문자 수
  • 93,418 hits
%d bloggers like this: