OMPM2001 Result: NEC NEC TX7 i9510 (1500MHz, Itanium 2) (tested by HLRS)

Benchmark	Reference Time	Base Runtime	Base Ratio	Peak Runtime	Peak Ratio
310.wupwise_m	6000	139	43072
312.swim_m	6000	195	30753
314.mgrid_m	7300	195	37496
316.applu_m	4000	61.7	64807
318.galgel_m	5100	307	16605
320.equake_m	2600	123	21095
324.apsi_m	3400	124	27430
326.gafort_m	8700	296	29411
328.fma3d_m	4600	193	23829
330.art_m	6400	116	55011
332.ammp_m	7000	372	18801
SPECompMbase2001	30649
	SPECompMpeak2001	--

Benchmark

Reference
Time

Base
Runtime

Base
Ratio

Peak
Runtime

Peak
Ratio

310.wupwise_m

6000

139

43072

312.swim_m

6000

195

30753

314.mgrid_m

7300

195

37496

316.applu_m

4000

61.7

64807

318.galgel_m

5100

307

16605

320.equake_m

2600

123

21095

324.apsi_m

3400

124

27430

326.gafort_m

8700

296

29411

328.fma3d_m

4600

193

23829

330.art_m

6400

116

55011

332.ammp_m

7000

372

18801

SPECompMbase2001

30649

SPECompMpeak2001

Hardware

Hardware Vendor:

NEC

Model Name:

NEC TX7 i9510 (1500MHz, Itanium 2)

CPU:

Intel Itanium 2

CPU MHz:

1500

FPU:

Integrated

CPU(s) enabled:

32 cores, 32 chips, 1 core/chip

CPU(s) orderable:

8-32

Primary Cache:

16KBI + 16KBD (on chip) per CPU

Secondary Cache:

256KB (on chip) per CPU

L3 Cache:

6.0MB (on chip) per CPU

Other Cache:

N/A

Memory:

512 GB

Disk Subsystem:

Seagate ST373307LC

Other Hardware:

None

Software

OpenMP Threads:

Parallel:

OpenMP

Operating System:

NEC Linux 3.4

Compiler:

Intel(R) Fortran Compiler for Linux 8.1 Build 20041123
Intel(R) C++ Compiler for Linux 8.1 Build 20041123

File System:

ext2 filesystem

System State:

Multi-user

Notes / Tuning Information
Tested by HLRS Baseline optimization flags: C programs: -openmp -O3 -IPF_fp_relaxed -ipo -ansi -ansi_alias (ONESTEP) Fortran programs: -openmp -O3 -IPF_fp_relaxed -ipo (ONESTEP) Portability Flags: 318.galgel_m: -FI -132 Extra Flags: 330.art_m: -DINTS_PER_CACHELINE=32 -DDBLS_PER_CACHELINE=16 Baseline user environment: OMP_NUM_THREADS 32 limit stacksize 64000 KMP_STACKSIZE 31M KMP_LIBRARY TURNAROUND OMP_DYNAMIC FALSE KMP_SCHEDULE static,balanced For all benchmarks threads were bound to cores using the following submit command: fplace -r -o 1,2 $command, This instructs fplace to skip placement of the lightweight OpenMP monitor thread, which is created prior to the slave threads. For a description of HLRS's compiler flags, portability flags, and system parameters used to generate this result, please refer to the HLRS-20050119-Linux-Intel8.1-IPF.txt file in the flags directory.

Notes / Tuning Information

Tested by HLRS
 Baseline optimization flags: 
   C programs:       -openmp -O3 -IPF_fp_relaxed -ipo -ansi -ansi_alias (ONESTEP)
   Fortran programs: -openmp -O3 -IPF_fp_relaxed -ipo (ONESTEP)
 
 Portability Flags:
   318.galgel_m: -FI -132
 
 Extra Flags:
   330.art_m: -DINTS_PER_CACHELINE=32 -DDBLS_PER_CACHELINE=16

 Baseline user environment:
   OMP_NUM_THREADS 32
   limit stacksize 64000
   KMP_STACKSIZE 31M
   KMP_LIBRARY TURNAROUND 
   OMP_DYNAMIC FALSE
   KMP_SCHEDULE static,balanced


 For all benchmarks threads were bound to cores using the following submit command:
 fplace -r -o 1,2 $command,
 This  instructs fplace to skip placement of the lightweight
 OpenMP monitor thread, which is created prior to the slave threads.
 
 For a description of HLRS's compiler flags, portability flags, and 
 system parameters used to generate this result, please refer to the
 HLRS-20050119-Linux-Intel8.1-IPF.txt file in the flags directory.

First published at SPEC.org on 03-Feb-2005

Generated on Thu Feb 3 11:09:42 2005 by SPEC OMP2001 HTML formatter v1.01