SPECint_rate95

IBM Corporation

RS/6000 SP-375MHz (32 nodes, 4 CPUs each)

This result has been submitted to SPEC, see the disclaimer before studying any results.

If you wish to print this information we highly recommend using the postscript reporting page.

Benchmark # and Name	Base # Copies	Base Run Time	Base SPEC Ratio	Peak # Copies	Peak Run Time	Peak SPEC Ratio
099.go	128	196	27096	128	200	26505
124.m88ksim	128	88.0	24878	128	89	24701
126.gcc	128	591	3316	128	678	2887
129.compress	128	96.5	21487	128	87	23873
130.li	128	106	20655	128	99	22185
132.ijpeg	128	129	21379	128	133	20719
134.perl	128	142	15378	128	110	19956
147.vortex	128	130	23916	128	121	25689
SPECint_rate_base95			17239
			SPECint_rate95			17920

Tester Information:

SPEC License #:	11
Tested By:	IBM, Poughkeepsie, NY
Test Date:	Feb-2000
Hardware Avail:	Feb-2000
Software Avail:	Feb-2000

Hardware Information:

Model Name:	RS/6000 SP-375MHz (32 nodes, 4 CPUs each)
CPU:	375 MHz Power3-II
FPU:	Integrated
Number of CPU(s):	128
Primary Cache:	32KB I + 64KB D (on chip)
Secondary Cache:	8MB unified (off chip) per CPU
Other Cache:	None
Memory:	4GB (per node)
Disk Subsystem:	2x9.1GB IBM Fast/Wide SCSI 7200 RPM (per node)
	16x9.1GB SSA 7200 RPM (GPFS)
Other Hardware:	SP Switch, Ethernet

Software Information:

Operating System:	AIX 4.3.3
Compiler:	IBM C/C++ 3.6.4.0
	memdbg 4.4.0.1
	FDPR
File System:	GPFS (distributed), AIX/JFS (local)
System State:	Multi-user

Notes:

Compatibility Flags: gcc -ma; perl -DI_TIME -DI_SYS_TIME
Base:   -O4 -Q=500 -qpdf1/pdf2 -lhmu
go:     -O -qarch=pwr3 -Q=500 -qansialias -qdatalocal -qassert=addr -qcompact -bnso -bI:/lib/syscalls.exp
m88:    -O3 -qarch=ppc -qipa=noobject -Q=500 -qansialias -qpdf1/pdf2
gcc:    -O -qarch=ppc -qdatalocal -qassert=addr -qinlglue -qpdf1/pdf2 -bnso -bI:/lib/syscalls.exp
comp:   -O -qarch=ppc -qipa=level=2:noobject -qdatalocal -Q=200 -qassert=allp -qunroll=5
li:     -O4 -qipa=noobject -qlibansi -qpdf1/pdf2; fdpr -R3
ijpeg:  -O4 -qipa=noobject -qunroll=4 -qpdf1/pdf2 -bnso -bI:/lib/syscalls.exp
perl:   -O4 -qipa=noobject -Q=200 -qinlglue -Dsetjmp=_setjmp -Dlongjmp=_longjmp -qansialias -qunroll=4
-qpdf1/pdf2 -bnso -bI:/lib/syscalls.exp /usr/ccs/lib/bmalloc.o
vort:   -O4 -qipa=level=2:noobject -Q=500 -qpdf1/pdf2 -lhm -lhu; fdpr -R2
submit command per node (x in v60n0x is node #):
rsh v60n0x "cd %dir%; PATH=/gpfs/wh2fs1/spec95:$PATH %command% && exit 0
NOTE: All benchmark related files are contained in a single GPFS filesystem.
The local filesystem contains the operating system and paging space.

Go To: [Home] [OSG] [CPU95] [Results]

This page was generated with rawformat (cd01-open).

First published at SPEC.org on 29-Feb-2000