SPEC® CFP2006 Result

Copyright 2006-2014 Standard Performance Evaluation Corporation

Hewlett-Packard Company

HP Integrity rx8640 (1.6GHz/24MB Dual-Core
Intel Itanium 2)

CPU2006 license: 03 Test date: Sep-2006
Test sponsor: Hewlett-Packard Company Hardware Availability: Sep-2006
Tested by: Hewlett-Packard Company Software Availability: Sep-2006
Benchmark results graph
Hardware
CPU Name: Dual-Core Intel Itanium 2 9050
CPU Characteristics: 1.6GHz/24MB, 533MHz FSB
CPU MHz: 1600
FPU: Integrated
CPU(s) enabled: 32 cores, 16 chips, 2 cores/chip
CPU(s) orderable: 1-16 chips
Primary Cache: 16 KB I + 16 KB D on chip per core
Secondary Cache: 1 MB I + 256 KB D on chip per core
L3 Cache: 12 MB I+D on chip per core
Other Cache: None
Memory: 128 GB (64x2GB DIMMs)
Disk Subsystem: 73GB 15K RPM SCSI
Other Hardware: None
Software
Operating System: HPUX11i-TCOE B.11.23.0609
Compiler: HP C/aC++ Developer's Bundle C.11.23.12
HP Fortran90 Compiler B.11.23.32
Auto Parallel: No
File System: vxfs
System State: Multi-user
Base Pointers: 32-bit
Peak Pointers: 32-bit
Other Software: None

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 32 1649 264 1643 265 1642 265 32 1649 264 1643 265 1642 265
416.gamess 32 2292 273 2253 278 2257 278 32 2158 290 2160 290 2156 291
433.milc 32 1613 182 1610 182 1609 183 32 1544 190 1544 190 1547 190
434.zeusmp 32 790 369 789 369 791 368 32 790 369 789 369 791 368
435.gromacs 32 519 441 515 444 515 443 32 457 500 458 499 455 503
436.cactusADM 32 457 838 456 838 457 836 32 457 838 456 838 457 836
437.leslie3d 32 1329 226 1330 226 1325 227 32 1329 226 1330 226 1325 227
444.namd 32 310 828 304 845 303 846 32 310 828 304 845 303 846
447.dealII 32 664 551 690 530 663 553 32 664 551 690 530 663 553
450.soplex 32 1436 186 1248 214 1248 214 32 1305 204 1085 246 1086 246
453.povray 32 597 285 594 287 595 286 32 495 344 488 349 488 349
454.calculix 32 603 438 588 449 587 449 32 603 438 588 449 587 449
459.GemsFDTD 32 2003 169 2006 169 2009 169 32 1933 176 1934 176 1929 176
465.tonto 32 1021 308 1020 309 1019 309 32 980 321 984 320 984 320
470.lbm 32 1075 409 1074 409 1073 410 32 1075 409 1074 409 1073 410
481.wrf 32 1014 353 1017 352 1007 355 32 1014 353 1017 352 1007 355
482.sphinx3 32 1039 600 1004 621 1002 622 32 1007 620 960 650 961 649

Operating System Notes

   The system had the September 2006 HP-UX 11i v2 Technical Computing
   Operating Environment (TCOE) and compilers installed, along with
   the following patches:

   PHSS_34858  linker + fdp cumulative patch
   PHSS_34853  Math Library Cumulative Patch
   PHSS_34854  Integrity Unwind Library
   PHSS_34855  HP C Compiler (A.06.12)
   PHSS_34856  aC++ Compiler (A.06.12)
   PHSS_34857  u2comp/be/plugin library patch
   PHSS_34395  FORTRAN I/O Library [libIO77]
   PHSS_34397  FORTRAN Intrinsics [libF90 B.11.23.17]
   PHSS_34399  Fortran Product Patch, v3.1 to v3.1.1
   PHKL_34020  Perfmon enhancements and Itanium Dual-Core

   The following kernel tunables were set, in addition to the defaults
   set by the Technical Computing OE:

   dbc_max_pct=20
   dbc_min_pct=20
   maxdsiz=3221225472
   maxssiz=401604608

Platform Notes

 The system was configured as a single partition with 4 cells and
 4 processors (8 cores) per cell.  Memory was configured as 50% local
 and 50% interleaved.

 The following config file entry was used to bind
 processes to cells using the HP-UX "mpsched" utility:
 submit = let "MYNUM=$SPECCOPYNUM" ; let "LDOM=\$MYNUM/8" ; mpsched -l \$LDOM $command

Base Compiler Invocation

C benchmarks:

 /opt/ansic/bin/cc -Ae 

C++ benchmarks:

 /opt/aCC/bin/aCC -Aa 

Fortran benchmarks:

 /opt/fortran90/bin/f90 

Benchmarks using both Fortran and C:

 /opt/ansic/bin/cc -Ae   /opt/fortran90/bin/f90 

Base Portability Flags

453.povray:  -DSPEC_CPU_NEED_INVHYP 
454.calculix:  -DSPEC_CPU_NOZMODIFIER 
481.wrf:  -DNOUNDERSCORE   +noppu 

Base Optimization Flags

C benchmarks:

 +Ofaster   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   -Wl,-N 

C++ benchmarks:

 +Ofaster   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   -Wl,-N 

Fortran benchmarks:

 +Ofaster   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   -Wl,-N 

Benchmarks using both Fortran and C:

 +Ofaster(-hp_cc)   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Ofaster(-hp_f90)   -Wl,-N 

Peak Compiler Invocation

C benchmarks:

 /opt/ansic/bin/cc -Ae 

C++ benchmarks:

 /opt/aCC/bin/aCC -Aa 

Fortran benchmarks:

 /opt/fortran90/bin/f90 

Benchmarks using both Fortran and C:

 /opt/ansic/bin/cc -Ae   /opt/fortran90/bin/f90 

Peak Portability Flags

453.povray:  -DSPEC_CPU_NEED_INVHYP 
454.calculix:  -DSPEC_CPU_NOZMODIFIER 
481.wrf:  -DNOUNDERSCORE   +noppu 

Peak Optimization Flags

C benchmarks:

433.milc:  +Oprofile=collect:all(pass 1)   +Oprofile=use(pass 2)   +Ofaster   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Onoparmsoverlap   -Wl,-N 
470.lbm:  basepeak = yes 
482.sphinx3:  +Oprofile=collect:all(pass 1)   +Oprofile=use(pass 2)   +Ofaster   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Onoparmsoverlap 

C++ benchmarks:

444.namd:  basepeak = yes 
447.dealII:  basepeak = yes 
450.soplex:  +Oprofile=collect:all(pass 1)   +Oprofile=use(pass 2)   +Ofaster   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Onoparmsoverlap   -Wl,-N 
453.povray:  +Oprofile=collect:all(pass 1)   +Oprofile=use(pass 2)   +Ofaster   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M 

Fortran benchmarks:

410.bwaves:  basepeak = yes 
416.gamess:  +Ofaster   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Odataprefetch=direct   -Wl,-N 
434.zeusmp:  basepeak = yes 
437.leslie3d:  basepeak = yes 
459.GemsFDTD:  +Oprofile=collect:all(pass 1)   +Oprofile=use(pass 2)   +Ofaster   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Odataprefetch=direct   -Wl,-N 
465.tonto:  +Oprofile=collect:all(pass 1)   +Oprofile=use(pass 2)   +Ofaster   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Odataprefetch=direct 

Benchmarks using both Fortran and C:

435.gromacs:  +Oprofile=collect:all(pass 1)   +Oprofile=use(pass 2)   +Ofaster(-hp_cc)   +Otype_safety=ansi   -Wl,-a,archive_shared   -Wl,+pd,64M   -Wl,+pi,64M   +Onoparmsoverlap   +Ofaster(-hp_f90) 
436.cactusADM:  basepeak = yes 
454.calculix:  basepeak = yes 
481.wrf:  basepeak = yes 

The flags file that was used to format this result can be browsed at
http://www.spec.org/cpu2006/flags/CPU2006_flags.20090715.06.html.

You can also download the XML flags source by saving the following link:
http://www.spec.org/cpu2006/flags/CPU2006_flags.20090715.06.xml.