CFP2000 Result: SGI SGI 2100 1X 350MHz R12k

Benchmark	Reference Time	Base Runtime	Base Ratio	Runtime	Ratio
168.wupwise	1600	644	248
171.swim	3100	1057	293
172.mgrid	1800	849	212
173.applu	2100	987	213
177.mesa	1400	576	243
178.galgel	2900	355	817
179.art	2600	300	867
183.equake	1300	655	199
187.facerec	1900	659	288
188.ammp	2200	693	318
189.lucas	2000	829	241
191.fma3d	2100	1204	174
200.sixtrack	1100	643	171
301.apsi	2600	1171	222
SPECfp_base2000	278
	SPECfp2000	--

Benchmark

Reference
Time

Base
Runtime

Base
Ratio

Runtime

Ratio

168.wupwise

1600

644

248

171.swim

3100

1057

293

172.mgrid

1800

849

212

173.applu

2100

987

213

177.mesa

1400

576

243

178.galgel

2900

355

817

179.art

2600

300

867

183.equake

1300

655

199

187.facerec

1900

659

288

188.ammp

2200

693

318

189.lucas

2000

829

241

191.fma3d

2100

1204

174

200.sixtrack

1100

643

171

301.apsi

2600

1171

222

SPECfp_base2000

278

SPECfp2000

Hardware

Hardware Vendor:

SGI

Model Name:

SGI 2100 1X 350MHz R12k

CPU:

R12000

CPU MHz:

350

FPU:

Integrated

CPU(s) enabled:

1 core, 1 chip, 1 core/chip

CPU(s) orderable:

1-8

Parallel:

Primary Cache:

32KBI + 32KBD on chip

Secondary Cache:

4MB(I+D) off chip

L3 Cache:

N/A

Other Cache:

N/A

Memory:

1 GB

Disk Subsystem:

1 x 18 GB FC, 2 x 18 GB FC (striped)

Other Hardware:

None

Software

Operating System:

IRIX 6.5.10f

Compiler:

MIPSpro 7.3.1.2m C, C++, Fortran90
SCSL 1.3 Math Library

File System:

xfs

System State:

Single-user

Notes / Tuning Information
Baseline optimization flags (for C benchmarks): PASS1 : -Ofast=ip27 -IPA:use_intrinsic -fb_create /tmp/SPEC2000/FBDIR_base/$(EXEBASE) PASS2 : -Ofast=ip27 -IPA:use_intrinsic -fb_opt /tmp/SPEC2000/FBDIR_base/$(EXEBASE) Baseline optimization flags (for Fortran benchmarks): -Ofast=ip27 -LNO:fusion=2 -OPT:fast_nint=OFF Portability Flags: 178.galgel: -fixedform Peak optimization flags: note: all occurances of (FEEDBACK) below means compiled with a two-step process: PASS1 = -fb_create /tmp/SPEC2000/FBDIR_peak/$(EXEBASE) PASS2 = -fb_opt /tmp/SPEC2000/FBDIR_peak/$(EXEBASE) 168.wupwise: -Ofast=ip27 -IPA:space=1000:linear=on:plimit=10000:callee_limit=5000 -INLINE:aggressive=on . -OPT:Olimit=0 -LNO:fusion=2:prefetch_ahead=5 171.swim: -Ofast=ip27 -LNO:cs2=8m 172.mgrid: -Ofast=ip27 -LNO:fusion=2 173.applu: -Ofast=ip27 -LNO:ou_max=5:ou_prod_max=10:prefetch=0:fusion=2 177.mesa: -Ofast=ip27 -OPT:goto=off -LNO:opt=0 (FEEDBACK) 178.galgel: -Ofast=ip27 -LNO:ou_max=7 -lscs (FEEDBACK) . RM_SOURCES = lapak.f90 179.art: -Ofast=ip27 -LNO:prefetch=0 -IPA:min_hot=15 (FEEDBACK) 183.equake: -Ofast=ip27 -LNO:prefetch=0 -TENV:X=4 -CG:ld_latency=7 -IPA:space=500 (FEEDBACK) 187.facerec: -Ofast=ip27 -fb_opt /tmp/SPEC2000/FBDIR_peak/$(EXEBASE) -LNO:fusion=2 (FEEDBACK) 188.ammp: -Ofast=ip27 -OPT:goto=off -IPA:space=500:plimit=900 (FEEDBACK) 189.lucas: -Ofast=ip27 -LNO:fusion=2:blocking=off -CG:ld_latency=4 -IPA:min_hot=8 (FEEDBACK) 191.fma3d: -Ofast=ip27 -bigp_off -LNO:prefetch=0 -OPT:goto=off:unroll_size=160:unroll_times_max=4 . -CG:ld_latency=2 (FEEDBACK) 200.sixtrack:= -Ofast=ip27 -IPA:maxdepth=2 -LNO:prefetch=0 (FEEDBACK) 301.apsi: -Ofast=ip27 -TENV:X=4 -LNO:prefetch=0:blocking=off -IPA:linear=on:use_intrinsic setenv PAGESIZE_DATA 4096 ; setenv PAGESIZE_TEXT 4096 ; setenv PAGESIZE_STACK 4096 systune -i ; percent_totalmem_4m_pages = 40 ; percent_totalmem_1m_pages = 7 systune -i ; percent_totalmem_256k_pages = 7 ; percent_totalmem_64k_pages = 7 limit stacksize 500000 The following is done before building each benchmark that requires (FEEDBACK): rm -rf /tmp/SPEC2000/FBDIR_peak/$baseexe ; mkdir -p /tmp/SPEC2000/FBDIR_peak/$baseexe

Notes / Tuning Information

Baseline optimization flags (for C benchmarks): 
   PASS1 : -Ofast=ip27 -IPA:use_intrinsic -fb_create /tmp/SPEC2000/FBDIR_base/$(EXEBASE)
   PASS2 : -Ofast=ip27 -IPA:use_intrinsic -fb_opt /tmp/SPEC2000/FBDIR_base/$(EXEBASE)
Baseline optimization flags (for Fortran benchmarks): -Ofast=ip27 -LNO:fusion=2 -OPT:fast_nint=OFF
Portability Flags:
   178.galgel: -fixedform
Peak optimization flags:
note:  all occurances of (FEEDBACK) below means compiled with a two-step process:
   PASS1 = -fb_create /tmp/SPEC2000/FBDIR_peak/$(EXEBASE)
   PASS2 = -fb_opt /tmp/SPEC2000/FBDIR_peak/$(EXEBASE)
168.wupwise: -Ofast=ip27 -IPA:space=1000:linear=on:plimit=10000:callee_limit=5000 -INLINE:aggressive=on 
.            -OPT:Olimit=0 -LNO:fusion=2:prefetch_ahead=5
171.swim: -Ofast=ip27 -LNO:cs2=8m
172.mgrid: -Ofast=ip27 -LNO:fusion=2
173.applu: -Ofast=ip27 -LNO:ou_max=5:ou_prod_max=10:prefetch=0:fusion=2
177.mesa: -Ofast=ip27 -OPT:goto=off -LNO:opt=0 (FEEDBACK)
178.galgel: -Ofast=ip27 -LNO:ou_max=7 -lscs (FEEDBACK)
.           RM_SOURCES = lapak.f90
179.art: -Ofast=ip27 -LNO:prefetch=0 -IPA:min_hot=15 (FEEDBACK)
183.equake: -Ofast=ip27 -LNO:prefetch=0 -TENV:X=4  -CG:ld_latency=7 -IPA:space=500 (FEEDBACK)
187.facerec: -Ofast=ip27 -fb_opt /tmp/SPEC2000/FBDIR_peak/$(EXEBASE) -LNO:fusion=2 (FEEDBACK)
188.ammp: -Ofast=ip27 -OPT:goto=off -IPA:space=500:plimit=900 (FEEDBACK)
189.lucas: -Ofast=ip27 -LNO:fusion=2:blocking=off -CG:ld_latency=4 -IPA:min_hot=8 (FEEDBACK)
191.fma3d: -Ofast=ip27 -bigp_off -LNO:prefetch=0 -OPT:goto=off:unroll_size=160:unroll_times_max=4 
.          -CG:ld_latency=2 (FEEDBACK)
200.sixtrack:= -Ofast=ip27 -IPA:maxdepth=2 -LNO:prefetch=0 (FEEDBACK)
301.apsi: -Ofast=ip27 -TENV:X=4 -LNO:prefetch=0:blocking=off -IPA:linear=on:use_intrinsic
   setenv PAGESIZE_DATA 4096 ; setenv PAGESIZE_TEXT 4096 ; setenv PAGESIZE_STACK 4096
   systune -i ; percent_totalmem_4m_pages = 40 ; percent_totalmem_1m_pages = 7
   systune -i ; percent_totalmem_256k_pages = 7 ; percent_totalmem_64k_pages = 7
   limit stacksize 500000
The following is done before building each benchmark that requires (FEEDBACK):
 rm -rf /tmp/SPEC2000/FBDIR_peak/$baseexe ; mkdir -p /tmp/SPEC2000/FBDIR_peak/$baseexe

First published at SPEC.org on 27-Mar-2001

Generated on Fri Mar 25 16:01:18 2005 by SPEC CPU2000 HTML formatter v1.01