| CPU2006 license: | 11 | Test date: | Apr-2015 |
|---|---|---|---|
| Test sponsor: | IBM Corporation | Hardware Availability: | Jun-2015 |
| Tested by: | IBM Corporation | Software Availability: | Mar-2015 |
| Hardware | |
|---|---|
| CPU Name: | POWER8 |
| CPU Characteristics: | Intelligent Energy Optimization enabled, up to 4.256 GHz |
| CPU MHz: | 4000 |
| FPU: | Integrated |
| CPU(s) enabled: | 192 cores, 16 chips, 12 cores/chip, 8 threads/core |
| CPU(s) orderable: | 4,8,12,16 Modules |
| Primary Cache: | 32 KB I + 64 KB D on chip per core |
| Secondary Cache: | 512 KB I+D on chip per core |
| L3 Cache: | 8 MB I+D on chip per core |
| Other Cache: | 16 MB I+D off chip per CDIMM |
| Memory: | 4 TB (128 x 32 GB CDIMMs) DDR3 1600 MHz |
| Disk Subsystem: | 16 x 300 GB 15K RPM SAS SFF-2 Raid5 |
| Other Hardware: | None |
| Software | |
|---|---|
| Operating System: | Red Hat Enterprise Linux Server release 7.1 (ppc64) kernel <3.10.0-229.1.2.el7.ppc64> |
| Compiler: | C/C++: Version 13.1 of IBM XL C/C++ for Linux; Fortran: Version 15.1 of IBM XL Fortran for Linux |
| Auto Parallel: | No |
| File System: | xfs |
| System State: | Run level 3 (multi-user) |
| Base Pointers: | 32-bit |
| Peak Pointers: | 32/64-bit |
| Other Software: | Post-Link Optimization for Linux on POWER, version 5.6.2 IBM Advance Toolchain 7.0-3 |
Results Table
| Benchmark | Base | Peak | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Copies | Seconds | Ratio | Seconds | Ratio | Seconds | Ratio | Copies | Seconds | Ratio | Seconds | Ratio | Seconds | Ratio | |
| Results appear in the order in which they were run. Bold underlined text indicates a median measurement. | ||||||||||||||
| 410.bwaves | 768 | 896 | 11600 | 895 | 11700 | 896 | 11700 | 192 | 198 | 13200 | 198 | 13200 | 213 | 12200 |
| 416.gamess | 768 | 1938 | 7760 | 1966 | 7650 | 2035 | 7390 | 768 | 1937 | 7760 | 1944 | 7730 | 1876 | 8020 |
| 433.milc | 768 | 476 | 14800 | 476 | 14800 | 476 | 14800 | 576 | 348 | 15200 | 350 | 15100 | 349 | 15200 |
| 434.zeusmp | 768 | 789 | 8850 | 782 | 8930 | 780 | 8960 | 768 | 692 | 10100 | 682 | 10300 | 680 | 10300 |
| 435.gromacs | 768 | 700 | 7830 | 697 | 7870 | 698 | 7860 | 768 | 576 | 9520 | 571 | 9600 | 574 | 9560 |
| 436.cactusADM | 768 | 794 | 11600 | 781 | 11800 | 783 | 11700 | 576 | 378 | 18200 | 380 | 18100 | 379 | 18100 |
| 437.leslie3d | 768 | 830 | 8700 | 834 | 8660 | 833 | 8670 | 384 | 354 | 10200 | 359 | 10100 | 358 | 10100 |
| 444.namd | 768 | 654 | 9420 | 632 | 9740 | 655 | 9410 | 1536 | 1243 | 9910 | 1229 | 10000 | 1233 | 9990 |
| 447.dealII | 768 | 506 | 17400 | 508 | 17300 | 505 | 17400 | 1152 | 701 | 18800 | 700 | 18800 | 699 | 18800 |
| 450.soplex | 768 | 1011 | 6330 | 659 | 9730 | 659 | 9720 | 768 | 973 | 6580 | 630 | 10200 | 630 | 10200 |
| 453.povray | 768 | 458 | 8910 | 459 | 8900 | 420 | 9730 | 1536 | 625 | 13100 | 624 | 13100 | 625 | 13100 |
| 454.calculix | 768 | 871 | 7280 | 857 | 7390 | 861 | 7360 | 768 | 859 | 7370 | 859 | 7380 | 860 | 7370 |
| 459.GemsFDTD | 768 | 1233 | 6610 | 1233 | 6610 | 1229 | 6630 | 768 | 1233 | 6610 | 1233 | 6610 | 1229 | 6630 |
| 465.tonto | 768 | 717 | 10500 | 725 | 10400 | 718 | 10500 | 1152 | 996 | 11400 | 1000 | 11300 | 998 | 11400 |
| 470.lbm | 768 | 1392 | 7580 | 1392 | 7580 | 1393 | 7580 | 192 | 172 | 15400 | 171 | 15400 | 174 | 15200 |
| 481.wrf | 768 | 769 | 11200 | 755 | 11400 | 752 | 11400 | 768 | 756 | 11400 | 739 | 11600 | 738 | 11600 |
| 482.sphinx3 | 768 | 1298 | 11500 | 1251 | 12000 | 1269 | 11800 | 384 | 607 | 12300 | 584 | 12800 | 582 | 12900 |
Peak Tuning Notes
410.bwaves fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 416.gamess fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 433.milc fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 434.zeusmp fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 435.gromacs fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 436.cactusADM fdpr options: -O4 -m power8 -A 2 -sls -dir -vrox 437.leslie3d fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 444.namd fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 447.dealII fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 453.povray fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 454.calculix fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 459.GemsFDTD fdpr options: -O4 -m power8 -A 2 -sls -dir -vrox 465.tonto fdpr options: -O4 -m power8 -A 2 -sls -dir -vrox 470.lbm fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 481.wrf fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox 482.sphinx3 fdpr options: -O4 -m power8 -A 2 -rcl 2 -sls -dir -vrox
Submit Notes
The config file option 'submit' was used to assign benchmark copy to specific kernel thread using the "numactl" command (see flags file for details).
Operating System Notes
ulimit -s (stack) set to unlimited 59520 16M large pages defined with echo 59520 > /proc/sys/vm/nr_hugepages echo 129600 > /proc/sys/vm/nr_overcommit_hugepages Transparent huge page enabled with echo always > /sys/kernel/mm/transparent_hugepage/enabled
General Notes
Environment variables set by runspec before the start of the run: HUGETLB_MORECORE = "yes" HUGETLB_VERBOSE = "0" TCMALLOC_MEMFS_MALLOC_PATH = "/dev/hugepages/" XLFRTEOPTS = "intrinthds=1" This result uses the post_setup and/or bench_post_setup to drop caches. SPEC has determined that although the effect may have been negligible for this run, future submissions will not be considered rule compliant if the post_setup actions drop caches (e.g. : "echo 3 > /proc/sys/vm/drop_caches").
Base Compiler Invocation
C benchmarks:
| /opt/ibm/xlC/13.1.0/bin/xlc_at -qlanglvl=extc99 |
C++ benchmarks:
| /opt/ibm/xlC/13.1.0/bin/xlC_at |
Fortran benchmarks:
| /opt/ibm/xlf/15.1.0/bin/xlf95_at |
Benchmarks using both Fortran and C:
| /opt/ibm/xlC/13.1.0/bin/xlc_at -qlanglvl=extc99 /opt/ibm/xlf/15.1.0/bin/xlf95_at |
Base Portability Flags
| 410.bwaves: | -qfixed |
| 416.gamess: | -qfixed |
| 434.zeusmp: | -qfixed |
| 435.gromacs: | -qfixed -qextname |
| 436.cactusADM: | -qfixed -qextname |
| 437.leslie3d: | -qfixed |
| 454.calculix: | -qfixed -qextname |
| 481.wrf: | -DNOUNDERSCORE |
| 482.sphinx3: | -qchars=signed |
Base Optimization Flags
C benchmarks:
| -qinline=40 -qipa=threads -qlargepage -O5 -qsimd=noauto -lhugetlbfs |
C++ benchmarks:
| -qinline=40 -qipa=threads -qlargepage -O5 -qrtti -lhugetlbfs |
Fortran benchmarks:
| -qipa=threads -qlargepage -O5 -qalias=nostd -lhugetlbfs |
Benchmarks using both Fortran and C:
| -qinline=40 -qipa=threads -qlargepage -O5 -qsimd=noauto -qalias=nostd -lhugetlbfs |
Base Other Flags
C benchmarks:
| -qipa=noobject -qsuppress=1500-036 |
C++ benchmarks:
| -qipa=noobject -qsuppress=1500-036 |
Fortran benchmarks:
| -qipa=noobject -qsuppress=1500-010 -qsuppress=cmpmsg -qsuppress=1500-036 |
Benchmarks using both Fortran and C:
| -qipa=noobject -qsuppress=1500-010 -qsuppress=cmpmsg -qsuppress=1500-036 |
Peak Compiler Invocation
C benchmarks:
| /opt/ibm/xlC/13.1.0/bin/xlc_at -qlanglvl=extc99 |
C++ benchmarks:
| /opt/ibm/xlC/13.1.0/bin/xlC_at |
Fortran benchmarks:
| /opt/ibm/xlf/15.1.0/bin/xlf95_at |
Benchmarks using both Fortran and C:
| /opt/ibm/xlC/13.1.0/bin/xlc_at -qlanglvl=extc99 /opt/ibm/xlf/15.1.0/bin/xlf95_at |
Peak Portability Flags
| 410.bwaves: | -qfixed |
| 416.gamess: | -qfixed |
| 434.zeusmp: | -qfixed |
| 435.gromacs: | -qfixed -qextname |
| 436.cactusADM: | -DSPEC_CPU_LP64 -qfixed -qextname |
| 437.leslie3d: | -qfixed |
| 454.calculix: | -qfixed -qextname |
| 481.wrf: | -DNOUNDERSCORE |
| 482.sphinx3: | -qchars=signed |
Peak Optimization Flags
C benchmarks:
| 433.milc: | -qinline=40 -qipa=threads -qlargepage -O5 -qsimd=noauto -qfdpr -lhugetlbfs -Wl,-q |
| 470.lbm: | -qinline=40 -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O5 -qsimd=noauto -q64 -qfdpr -lhugetlbfs -Wl,-q |
| 482.sphinx3: | -qinline=40 -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O5 -qsimd=noauto -qfdpr -lhugetlbfs -Wl,-q |
C++ benchmarks:
| 444.namd: | -qinline=40 -qipa=threads -qlargepage -O4 -qfdpr -lhugetlbfs -Wl,-q |
| 447.dealII: | -qinline=40 -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O4 -qfdpr -qrtti -lhugetlbfs -Wl,-q |
| 450.soplex: | -qinline=40 -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O3 -qarch=auto -qtune=auto -qsimd -qnoprefetch -lhugetlbfs |
| 453.povray: | -qinline=40 -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O3 -qarch=auto -qtune=auto -qprefetch=dscr=0x93 -qfdpr -lhugetlbfs -Wl,-q |
Fortran benchmarks:
| 410.bwaves: | -qipa=threads -qlargepage -O5 -qsimd=noauto -qfdpr -qsmallstack=dynlenonheap -lhugetlbfs -Wl,-q |
| 416.gamess: | -qipa=threads -qlargepage -O5 -qsimd=noauto -qprefetch=dscr=0x54 -qipa=partition=large -qfdpr -qalias=nostd -lhugetlbfs -Wl,-q |
| 434.zeusmp: | -qipa=threads -qlargepage -O4 -qsimd=noauto -q64 -qfdpr -qxlf90=nosignedzero -lhugetlbfs -Wl,-q |
| 437.leslie3d: | -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O5 -q64 -qfdpr -lhugetlbfs -Wl,-q -B/opt/at7.0/share/libhugetlbfs/ -tl -Wl,--hugetlbfs-align |
| 459.GemsFDTD: | basepeak = yes |
| 465.tonto: | -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O5 -q64 -qipa=partition=large -qfdpr -lhugetlbfs -Wl,-q |
Benchmarks using both Fortran and C:
| 435.gromacs: | -qinline=40 -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O4 -qipa=partition=large -qfdpr -lhugetlbfs -Wl,-q |
| 436.cactusADM: | -qinline=40 -qipa=threads -qpdf1(pass 1) -qpdf2(pass 2) -qlargepage -O4 -qarch=pwr7 -qtune=pwr7 -qipa=partition=large -q64 -qfdpr -lhugetlbfs -Wl,-q |
| 454.calculix: | -qinline=40 -qipa=threads -O5 -qsimd=noauto -qfdpr -lhugetlbfs -Wl,-q |
| 481.wrf: | -qinline=40 -qipa=threads -qlargepage -O5 -qipa=partition=large -qfdpr -lhugetlbfs -Wl,-q |
Peak Other Flags
C benchmarks (except as noted below):
| -qsuppress=1586-476(pass 2) -qipa=noobject -qsuppress=1500-036 | |
| 433.milc: | -qipa=noobject -qsuppress=1500-036 |
C++ benchmarks (except as noted below):
| -qsuppress=1586-476(pass 2) -qipa=noobject -qsuppress=1500-036 | |
| 444.namd: | -qipa=noobject -qsuppress=1500-036 |