SPEC OMPM2001 Summary Hewlett-Packard Company AlphaServer GS1280 Model M16 Wed Dec 4 05:33:13 2002 SPEC License #HPG2116 Tester: Hewlett-Packard Company Test date: Dec-2002 Test Site: Hewlett-Packard Company Hardware availability: Jan-2003 Software availability: Jan-2003 Base Base Base Peak Peak Peak Benchmarks Ref Time Run Time Ratio Ref Time Run Time Ratio ------------- -------- -------- -------- -------- -------- -------- 310.wupwise_m 6000 1191 5036* 6000 711 8440 310.wupwise_m 6000 1230 4877 6000 711 8435* 310.wupwise_m 6000 1191 5037 6000 712 8430 312.swim_m 6000 279 21529* 6000 279 21536 312.swim_m 6000 279 21542 6000 279 21521 312.swim_m 6000 279 21516 6000 279 21521* 314.mgrid_m 7300 912 8001 7300 894 8168 314.mgrid_m 7300 913 8000 7300 895 8152 314.mgrid_m 7300 912 8000* 7300 895 8156* 316.applu_m 4000 394 10147 4000 378 10591* 316.applu_m 4000 394 10158* 4000 377 10613 316.applu_m 4000 394 10161 4000 381 10504 318.galgel_m 5100 593 8595 5100 606 8419 318.galgel_m 5100 593 8597* 5100 606 8411* 318.galgel_m 5100 593 8601 5100 606 8410 320.equake_m 2600 586 4435 2600 580 4485* 320.equake_m 2600 585 4448* 2600 579 4489 320.equake_m 2600 584 4450 2600 583 4459 324.apsi_m 3400 511 6652 3400 522 6516 324.apsi_m 3400 519 6546 3400 521 6529* 324.apsi_m 3400 519 6549* 3400 520 6539 326.gafort_m 8700 1825 4768 8700 1700 5118* 326.gafort_m 8700 1826 4765* 8700 1700 5118 326.gafort_m 8700 1826 4764 8700 1700 5119 328.fma3d_m 4600 990 4646 4600 942 4884* 328.fma3d_m 4600 989 4651 4600 942 4885 328.fma3d_m 4600 990 4649* 4600 942 4884 330.art_m 6400 2411 2655 6400 1207 5302 330.art_m 6400 2401 2665* 6400 1215 5266* 330.art_m 6400 2397 2670 6400 1222 5237 332.ammp_m 7000 2394 2923 7000 2378 2944 332.ammp_m 7000 2388 2931 7000 2380 2942* 332.ammp_m 7000 2394 2924* 7000 2380 2941 ======================================================================== 310.wupwise_m 6000 1191 5036* 6000 711 8435* 312.swim_m 6000 279 21529* 6000 279 21521* 314.mgrid_m 7300 912 8000* 7300 895 8156* 316.applu_m 4000 394 10158* 4000 378 10591* 318.galgel_m 5100 593 8597* 5100 606 8411* 320.equake_m 2600 585 4448* 2600 580 4485* 324.apsi_m 3400 519 6549* 3400 521 6529* 326.gafort_m 8700 1826 4765* 8700 1700 5118* 328.fma3d_m 4600 990 4649* 4600 942 4884* 330.art_m 6400 2401 2665* 6400 1215 5266* 332.ammp_m 7000 2394 2924* 7000 2380 2942* SPECompMbase2001 6027 SPECompMpeak2001 6824 HARDWARE -------- Hardware Vendor: Hewlett-Packard Company Model Name: AlphaServer GS1280 Model M16 CPU: Alpha 21364 CPU MHz: 1150 FPU: Integrated CPU(s) enabled: 16 CPU(s) orderable: 2 to 16 by 2 Primary Cache: 64KB(I)+64KB(D) on chip Secondary Cache: 1.75MB on chip per CPU L3 Cache: None Other Cache: None Memory: 64GB Disk Subsystem: 36GB 10k rpm ufs Other Hardware: None SOFTWARE -------- OpenMP Threads: 4 Parallel: OpenMP Operating System: Tru64 UNIX V5.1B (Rev 2650) + IPK Compiler: Compaq Fortran X5.5-2602-48C8L Compaq C V6.5-011-48C5K BETA DCPI - 3.9.6 (20020307 1815) SPIKE V5.2 (503DTK) GEM 48C5S LIBMLD 2.4 DATE APR 9 2002 File System: ufs System State: Multi-user NOTES ----- Base: cc -arch ev7 -fast -omp -O4 f90 -arch ev7 -fast -omp -O5 Peak: All use -arch ev7 -omp ONESTEP Individual benchmark tuning: 310.wupwise_m: f90 -call_shared -inline all -unroll 12 -align commons +PFB 312.swim_m: f90 -arch ev7 -fast -omp -O5 314.mgrid_m: f90 -O5 -transform_loops -tune ev7 -unroll 8 +PFB 316.applu_m: f90 -fast -O5 -unroll 14 +PFB 318.galgel_m: f90 -fast -O5 -unroll 5 -extend_source +PFB 320.equake_m: cc -fast -call_shared -O4 -ldensemalloc -assume restricted_pointers -inline speed -unroll 13 +PFB 324.apsi_m: f90 -O5 -transform_loops -unroll 8 +PFB 326.gafort_m: f90 -fast -O5 -arch ev67 -tune ev67 328.fma3d: f90 -O4 -transform_loops 330.art_m: cc -assume whole_program -ldensemalloc -call_shared -assume restricted_pointers -fast -O4 -unroll 16 -inline none +PFB 332.ammp_m: cc -O4 -ifo -assume nomath_errno -assume trusted_short_alignment -fp_reorder -readonly_strings -ldensemalloc -assume restricted_pointers -unroll 9 Portability: 318.galgel_m: -exend_source used in base and peak. Peak Source: Available as SPEC OMPM source:ompm2001-srcl32bit-20020831.tar.gz For 310.wupwise_m, 312.swim_m, 314.mgrid_m, 316.applu_m, 320.equake_m 324.apsi-m, 326.gafort_m, and 328.fma3d_m Available as SPEC OMPM source: ompm2001-isoc-20020619.tar.gz For 330.art_m User Environment: MP_STACK_SIZE = 10000000 OMP_NUM_THREADS=4 PTHREAD_CONFIG=feature=def-scs,d4-scs 4 CPU processor set used (man processor_sets) Restricts memory accessible to 16 GB Cpus 0,1,2,3 used. System tunables: 4Mb pages used - vm_bigpg_enabled set in sysconfigtab Description of +PFB: Prefetches are improved by the post-link-time optimizer Spike, using feedback from a training run. These commands are used (in phase "fdo_post_makeN"): rm -rf db mkdir db dcpid -no_palcode ./db A training run is carried out (in phase "fdo_runN"), and then these commands (in phase "fdo_postN"): dcpiquit dcpi2bb -make-bbdb -pm all -counts -conf_low -db ./db ${baseexe} spike ${baseexe} -feedback ${baseexe} -o newexe rm ${baseexe} mv newexe ${baseexe} Information on UNIX V5.1 Patches can be found at http://www.support.compaq.com/patches Information on DCPI and SPIKE can be found at http://www.tru64unix.compaq.com/dtk ----------------------------------------------------------------------------- For questions about this result, please contact the tester. For other inquiries, please contact webmaster@spec.org. Copyright 1999-2002 Standard Performance Evaluation Corporation Generated on Wed Dec 25 14:36:38 2002 by SPEC OMP2001 ASCII formatter v2.1