|
|
OMPL2001 Result Copyright © 1999-2007 Standard Performance Evaluation Corporation |
|
IBM Corporation IBM eServer p5 595 (1900 MHz, 64 CPU) |
|
| SPEC license # HPG0005 | Tested by: IBM | Test site: Austin, TX | Test date: Jan-2005 | Hardware Avail: Nov-2005 | Software Avail: Dec-2004 |
| Benchmark | Reference Time |
Base Runtime |
Base Ratio |
Peak Runtime |
Peak Ratio |
![]() |
|---|---|---|---|---|---|---|
| 311.wupwise_l | 9200 | 168 | 875814 | 168 | 875814 | |
| 313.swim_l | 12500 | 307 | 652359 | 275 | 726223 | |
| 315.mgrid_l | 13500 | 415 | 521055 | 415 | 521055 | |
| 317.applu_l | 13500 | 265 | 816109 | 243 | 889842 | |
| 321.equake_l | 13000 | 797 | 261113 | 797 | 261113 | |
| 325.apsi_l | 10500 | 330 | 509199 | 330 | 509199 | |
| 327.gafort_l | 11000 | 272 | 646229 | 272 | 646229 | |
| 329.fma3d_l | 23500 | 752 | 499790 | 752 | 499790 | |
| 331.art_l | 25000 | 305 | 1311513 | 179 | 2229235 | |
| SPECompLbase2001 | 620741 | |||||
| SPECompLpeak2001 | 672757 | |||||
|
|
||||||||||||||||||||||||||||||||||||||||||||
| Notes / Tuning Information |
|---|
Tested by IBM
Portability Flags & Environment Variables
-qfixed used in: 311.wupwise_l, 313.swim_l, 315.mgrid_l, 317.applu_l, 325.apsi_l
-qfixed=80 used in: 319.galgel_l
-qsuffix=f=f90 used in: 319.galgel_l, 327.gafort_l, 329.fma3d_l
Base Flags
C: -q64 -O5 -qalign=natural -lmass -qsmp=omp
FORTRAN: -q64 -O5 -lmass -qsmp=omp
Base & Peak User Environment:
OMP_NUM_THREADS=128
OMP_DYNAMIC=FALSE
XLSMPOPTS=SPINS=0:YIELDS=0:STACK=8000000:SCHEDULE=STATIC:STARTPROC=0:STRIDE=1
MALLOCMULTIHEAP=1
MEMORY_AFFINITY=MCM
Peak Flags:
-qsmp=omp used in all cases
311.wupwise_l: basepeak=1
313.swim_l: -q64 -O3 -qarch=pwr3 -qtune=pwr3
315.mgrid_l: basepeak=1
XLSMPOPTS=SPINS=0:YIELDS=0:STACK=8000000:SCHEDULE=STATIC:STARTPROC=0:STRIDE=2
OMP_NUM_THREADS=64
317.applu_l: -q64 -O3 -qarch=pwr4 -qtune=pwr4
XLSMPOPTS=SPINS=0:YIELDS=0:STACK=8000000:SCHEDULE=STATIC:STARTPROC=0:STRIDE=2
OMP_NUM_THREADS=64
319.galgel_l: -q64 -O5 -qhot=arraypad -qipa=noobject -qipa=partition=large -qmaxmem=-1
321.equake_l: basepeak=1
XLSMPOPTS=SPINS=0:YIELDS=0:STACK=8000000:SCHEDULE=STATIC:STARTPROC=0:STRIDE=2
OMP_NUM_THREADS=64
325.apsi_l: basepeak=1
327.gafort_l: basepeak=1
329.fma3d_l: basepeak=1
331.art_l: -q64 -O5 -qalign=natural -lmass -blpdata
EXTRA_CFLAGS= -DINTS_PER_CACHELINE=32 -DDBLS_PER_CACHELINE=16
Alternate sources:
Add critical region around update of linked list in parallel loop.
Required src.alt available as ompm-purdue1-20040324.tar.gz
Used for 331.art_l, base and peak.
APAR IY62267 was applied to AIX 5L V5.3 to achieve Mantainence Level 1.
SMT: Acronym for "Simultaneous Multi-Threading". A processor technology that allows
the simultaneous execution of multiple thread contexts within a single processor
core. (Enabled by default)
MCM: Acronym for "Multi-Chip Module" (four dual-core processor chips + four L3-cache chips)
ESSL: Engineering and Scientific Subroutine Library
SUT: Acronym for "System Under Test"
C: IBM XL C for AIX invoked as xlc_r
Fortran 90: IBM XL Fortran for AIX invoked as xlf90_r
ulimits set to unlimited.
Large page mode and memory affinity were set as follows:
vmo -r -o lgpg_regions=4096 -o lgpg_size=16777216 -o memory_affinity=1
chuser capabilities=CAP_BYPASS_RAC_VMM,CAP_PROPAGATE $USER
shutdown -r
|
First published at SPEC.org on 03-Mar-2005
Generated on Thu Jun 28 17:59:48 2007 by SPEC OMP2001 HTML formatter v1.01