| CPU2006 license: | 11 | Test date: | May-2007 |
|---|---|---|---|
| Test sponsor: | IBM Corporation | Hardware Availability: | Jun-2007 |
| Tested by: | IBM Corporation | Software Availability: | Jun-2007 |
| Hardware | |
|---|---|
| CPU Name: | POWER6 |
| CPU Characteristics: | |
| CPU MHz: | 4700 |
| FPU: | Integrated |
| CPU(s) enabled: | 16 cores, 8 chips, 2 cores/chip, 2 threads/core |
| CPU(s) orderable: | 2,4,8,12,16 cores |
| Primary Cache: | 64 KB I + 64 KB D on chip per core |
| Secondary Cache: | 4 MB I+D on chip per core |
| L3 Cache: | 32 MB I+D off chip per chip |
| Other Cache: | None |
| Memory: | 128 GB (64x2 GB) DDR2 667 MHz |
| Disk Subsystem: | 2x73 GB 2x146 GB SAS 15K RPM |
| Other Hardware: | None |
| Software | |
|---|---|
| Operating System: | IBM AIX 5L V5.3 |
| Compiler: | XL C/C++ Enterprise Edition Version 9.0 for AIX |
| Auto Parallel: | No |
| File System: | AIX/JFS2 |
| System State: | Multi-user |
| Base Pointers: | 32-bit |
| Peak Pointers: | 32/64-bit |
| Other Software: | -- |
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. | ||||||||||||||
| 400.perlbench | 32 | 1146 | 273 | 1132 | 276 | 1132 | 276 | 32 | 954 | 328 | 954 | 328 | 928 | 337 |
| 401.bzip2 | 32 | 1003 | 308 | 1000 | 309 | 962 | 321 | 32 | 801 | 386 | 802 | 385 | 761 | 406 |
| 403.gcc | 32 | 659 | 391 | 662 | 389 | 662 | 389 | 32 | 603 | 427 | 600 | 430 | 594 | 434 |
| 429.mcf | 32 | 410 | 712 | 409 | 713 | 409 | 714 | 32 | 410 | 712 | 409 | 713 | 409 | 714 |
| 445.gobmk | 32 | 862 | 389 | 862 | 389 | 863 | 389 | 32 | 756 | 444 | 755 | 444 | 756 | 444 |
| 456.hmmer | 32 | 896 | 333 | 895 | 334 | 892 | 335 | 32 | 731 | 408 | 732 | 408 | 730 | 409 |
| 458.sjeng | 32 | 1126 | 344 | 1127 | 343 | 1127 | 344 | 32 | 1018 | 380 | 1019 | 380 | 1019 | 380 |
| 462.libquantum | 32 | 781 | 849 | 782 | 848 | 783 | 847 | 16 | 255 | 1300 | 255 | 1300 | 256 | 1300 |
| 464.h264ref | 32 | 1033 | 686 | 1034 | 685 | 1034 | 685 | 32 | 968 | 731 | 967 | 732 | 967 | 732 |
| 471.omnetpp | 32 | 640 | 312 | 639 | 313 | 639 | 313 | 32 | 549 | 365 | 548 | 365 | 548 | 365 |
| 473.astar | 32 | 719 | 312 | 717 | 313 | 718 | 313 | 32 | 639 | 351 | 637 | 353 | 636 | 353 |
| 483.xalancbmk | 32 | 595 | 371 | 591 | 374 | 584 | 378 | 32 | 505 | 438 | 504 | 438 | 504 | 438 |
General Notes
AIX 5L V5.3 updated with the 5300-06 Technology Level. See flags file for details on following settings. all ulimits set to unlimited Envionment variables set before executing benchmarks: MALLOCOPTIONS=pool MEMORY_AFFINITY=MCM XLFRTEOPTS=intrinthds=1 System set to "Enhanced" mode when defining partition on HMC 6144 pages of size 16M defined on systems with vmo command fdpr binary optimization tool used for peak versions of 401.bzip2 403.gcc 429.mcf 456.hmmer 462.libquantum 473.astar submit used to bind benchmark to a processor using "bindprocessor" The "IBM System p 570" and "IBM System i 570" are electronically equivalent. The results have been measured on the "IBM System p 570" model.
Base Compiler Invocation
C benchmarks:
| /usr/vac/bin/xlc -qlanglvl=extc99 |
C++ benchmarks:
| /usr/vacpp/bin/xlC |
Base Portability Flags
| 400.perlbench: | -DSPEC_CPU_AIX |
| 462.libquantum: | -DSPEC_CPU_AIX |
| 464.h264ref: | -DSPEC_CPU_AIX -qchars=signed |
| 483.xalancbmk: | -DSPEC_CPU_AIX |
Base Optimization Flags
C benchmarks:
| -bmaxdata:0x50000000 -O5 -qlargepage -D_ILS_MACROS -qalias=noansi -qalloca -blpdata |
C++ benchmarks:
| -bmaxdata:0x20000000 -O5 -qlargepage -D_ILS_MACROS -qrtti=all -blpdata |
Base Other Flags
C benchmarks:
| -qipa=noobject -qipa=threads -qsuppress=1500-036 |
C++ benchmarks:
| -qipa=noobject -qipa=threads -qsuppress=1500-036 |
Peak Compiler Invocation
C benchmarks:
| /usr/vac/bin/xlc -qlanglvl=extc99 |
C++ benchmarks:
| /usr/vacpp/bin/xlC |
Peak Portability Flags
| 400.perlbench: | -DSPEC_CPU_AIX |
| 403.gcc: | -DSPEC_CPU_LP64 |
| 462.libquantum: | -DSPEC_CPU_AIX |
| 464.h264ref: | -DSPEC_CPU_AIX -qchars=signed |
| 483.xalancbmk: | -DSPEC_CPU_AIX |
Peak Optimization Flags
C benchmarks:
| 400.perlbench: | -bmaxdata:0x50000000 -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -qenablevmx -qvecnvol -D_ILS_MACROS -qalias=noansi -blpdata |
| 401.bzip2: | -bmaxdata:0x4ffffffc -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -qenablevmx -qvecnvol -D_ILS_MACROS -blpdata |
| 403.gcc: | -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -D_ILS_MACROS -qalloca -q64 -blpdata |
| 429.mcf: | basepeak = yes |
| 445.gobmk: | -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -qenablevmx -qvecnvol -D_ILS_MACROS -blpdata |
| 456.hmmer: | -O5 -qlargepage -D_ILS_MACROS -blpdata |
| 458.sjeng: | -qpdf1(pass 1) -qpdf2(pass 2) -O4 -qlargepage -qenablevmx -qvecnvol -D_ILS_MACROS -blpdata |
| 462.libquantum: | -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -qenablevmx -qvecnvol -D_ILS_MACROS -q64 -blpdata |
| 464.h264ref: | -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -D_ILS_MACROS -blpdata |
C++ benchmarks:
| 471.omnetpp: | -bmaxdata:0x20000000 -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -qenablevmx -qvecnvol -D_ILS_MACROS -qalign=natural -qrtti=all -qinlglue -blpdata |
| 473.astar: | -bmaxdata:0x20000000 -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -qenablevmx -qvecnvol -D_ILS_MACROS -blpdata |
| 483.xalancbmk: | -bmaxdata:0x20000000 -qpdf1(pass 1) -qpdf2(pass 2) -O5 -qlargepage -D_ILS_MACROS -qinlglue -D__IBM_FAST_VECTOR -blpdata |