| CPU2006 license: | 19 | Test date: | Jun-2008 |
|---|---|---|---|
| Test sponsor: | Fujitsu Limited | Hardware Availability: | Jul-2008 |
| Tested by: | Sun Microsystems | Software Availability: | Jul-2008 |
| Hardware | |
|---|---|
| CPU Name: | SPARC64 VII |
| CPU Characteristics: | |
| CPU MHz: | 2520 |
| FPU: | Integrated |
| CPU(s) enabled: | 64 cores, 16 chips, 4 cores/chip, 2 threads/core |
| CPU(s) orderable: | 1 to 4 CMUs; each CMU contains 2 or 4 chips |
| Primary Cache: | 64 KB I + 64 KB D on chip per core |
| Secondary Cache: | 6 MB I+D on chip per chip |
| L3 Cache: | None |
| Other Cache: | None |
| Memory: | 256 GB (128 x 2 GB) |
| Disk Subsystem: | Fujitsu 73 GB 10000 RPM SAS |
| Other Hardware: | None |
| Software | |
|---|---|
| Operating System: | Solaris 10 5/08 with Patch 137111-03 |
| Compiler: | Sun Studio 12 with patches 124867-06, 124861-07, 124863-05, 127000-05 (see patch information below) |
| Auto Parallel: | Yes |
| File System: | ufs |
| System State: | Default |
| Base Pointers: | 32-bit |
| Peak Pointers: | 32-bit |
| Other Software: | None |
Results Table
| Benchmark | Base | Peak | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Seconds | Ratio | Seconds | Ratio | Seconds | Ratio | 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 | 35.2 | 386 | 36.8 | 369 | 36.6 | 371 | 27.5 | 494 | 27.5 | 495 | 27.3 | 497 |
| 416.gamess | 1868 | 10.5 | 1869 | 10.5 | 1869 | 10.5 | 1814 | 10.8 | 1816 | 10.8 | 1811 | 10.8 |
| 433.milc | 1032 | 8.89 | 1141 | 8.05 | 1139 | 8.06 | 699 | 13.1 | 684 | 13.4 | 698 | 13.1 |
| 434.zeusmp | 292 | 31.2 | 292 | 31.1 | 292 | 31.2 | 292 | 31.2 | 292 | 31.1 | 292 | 31.2 |
| 435.gromacs | 768 | 9.30 | 767 | 9.30 | 768 | 9.30 | 644 | 11.1 | 644 | 11.1 | 644 | 11.1 |
| 436.cactusADM | 53.6 | 223 | 52.1 | 230 | 52.7 | 227 | 52.5 | 228 | 51.9 | 230 | 51.6 | 231 |
| 437.leslie3d | 216 | 43.5 | 219 | 43.0 | 216 | 43.4 | 216 | 43.5 | 219 | 43.0 | 216 | 43.4 |
| 444.namd | 704 | 11.4 | 704 | 11.4 | 706 | 11.4 | 699 | 11.5 | 699 | 11.5 | 701 | 11.4 |
| 447.dealII | 846 | 13.5 | 846 | 13.5 | 845 | 13.5 | 846 | 13.5 | 846 | 13.5 | 845 | 13.5 |
| 450.soplex | 973 | 8.57 | 973 | 8.57 | 972 | 8.58 | 756 | 11.0 | 754 | 11.1 | 755 | 11.1 |
| 453.povray | 543 | 9.79 | 542 | 9.82 | 542 | 9.81 | 389 | 13.7 | 388 | 13.7 | 388 | 13.7 |
| 454.calculix | 654 | 12.6 | 654 | 12.6 | 654 | 12.6 | 630 | 13.1 | 631 | 13.1 | 630 | 13.1 |
| 459.GemsFDTD | 143 | 74.2 | 142 | 74.7 | 142 | 74.5 | 143 | 74.2 | 142 | 74.7 | 142 | 74.5 |
| 465.tonto | 1111 | 8.85 | 1112 | 8.85 | 1112 | 8.85 | 725 | 13.6 | 724 | 13.6 | 728 | 13.5 |
| 470.lbm | 40.2 | 342 | 40.2 | 342 | 40.2 | 342 | 40.2 | 342 | 40.2 | 342 | 40.2 | 342 |
| 481.wrf | 679 | 16.5 | 679 | 16.5 | 679 | 16.5 | 679 | 16.5 | 679 | 16.5 | 679 | 16.5 |
| 482.sphinx3 | 1862 | 10.5 | 1862 | 10.5 | 1871 | 10.4 | 1299 | 15.0 | 1278 | 15.3 | 1301 | 15.0 |
Compiler Invocation Notes
Sun Studio compiler patches are available at
http://developers.sun.com/sunstudio/downloads/patches/ss12_patches.jsp
Operating System Notes
Environment Variable Settings:
The maximum number of threads a program can create was set with:
OMP_NUM_THREADS=64
Program threads were bound to processors with:
SUNW_MP_PROCBIND=" 1 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70
72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 102 104
104 106 108 110 112 114 116 118 120 122 124 126 "
Behavior of parallel threads was set with:
SUNW_MP_THR_IDLE=SPIN
SPIN specifies that an idle thread should spin while waiting at barrier
or waiting for new parallel regions to work on.
System Tunables (/etc/system parameters):
tune_t_fsflushr=10
Controls how many seconds elapse between runs of the
page flush daemon, fsflush.
autoup=300
Causes pages older than the listed number of seconds to
be written by fsflush.
bufhwm=3000
Memory byte limit for caching I/O buffers
segmap_percent=3
Set maximum percent memory for file system cache
lpg_alloc_prefer=1
Set lgroup page allocation to strongly prefer local pages
Other System Settings:
The webconsole service was turned off using
svcadm disable webconsole
Platform Notes
Memory is 8-way interleaved by filling all slots with the same capacity DIMMs. This result is measured on a Sun SPARC Enterprise M8000 Server. Note that the Sun SPARC Enterprise M8000 and Fujitsu SPARC Enterprise M8000 are electrically equivalent.
Base Compiler Invocation
C benchmarks:
| cc |
C++ benchmarks:
| CC |
Fortran benchmarks:
| f90 |
Benchmarks using both Fortran and C:
| cc f90 |
Base Optimization Flags
C benchmarks:
| -fast -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xautopar -xreduction -xprefetch_level=3 -xprefetch_auto_type=indirect_array_access |
C++ benchmarks:
| -library=stlport4 -fast -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xautopar -xreduction -xprefetch_level=2 -xalias_level=compatible |
Fortran benchmarks:
| -fast -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xautopar -xreduction -xprefetch_level=2 |
Benchmarks using both Fortran and C:
| -fast(cc) -fast(f90) -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xautopar -xreduction -xprefetch_level=3 -xprefetch_auto_type=indirect_array_access -xprefetch_level=2 |
Base Other Flags
C benchmarks:
| -xjobs=64 |
C++ benchmarks:
| -xjobs=64 |
Fortran benchmarks:
| -xjobs=64 |
Benchmarks using both Fortran and C:
| -xjobs=64 |
Peak Compiler Invocation
C benchmarks:
| cc |
C++ benchmarks:
| CC |
Fortran benchmarks:
| f90 |
Benchmarks using both Fortran and C:
| cc f90 |
Peak Optimization Flags
C benchmarks:
| 433.milc: | -fast -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xalias_level=std -xprefetch_level=3 -xprefetch_auto_type=indirect_array_access -xalias_level=strong -xprefetch_level=2 |
| 470.lbm: | basepeak = yes |
| 482.sphinx3: | -xprofile=collect:./feedback(pass 1) -xprofile=use:./feedback(pass 2) -fast -xipo=2 -fma=fused -xalias_level=std -xautopar -xreduction |
C++ benchmarks:
Fortran benchmarks:
| 410.bwaves: | -fast -xipo=2 -fma=fused -xpagesize=512K -xprefetch=latx:2 -xprefetch_level=2 -xautopar -xreduction |
| 416.gamess: | -fast -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xprefetch_level=2 |
| 434.zeusmp: | basepeak = yes |
| 437.leslie3d: | basepeak = yes |
| 459.GemsFDTD: | basepeak = yes |
| 465.tonto: | -xprofile=collect:./feedback(pass 1) -xprofile=use:./feedback(pass 2) -fast -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xarch=v8plusa -xprefetch=latx:12 -lfast |
Benchmarks using both Fortran and C:
| 435.gromacs: | -xprofile=collect:./feedback(pass 1) -xprofile=use:./feedback(pass 2) -fast(cc) -fast(f90) -xipo=2 -fma=fused -xalias_level=std |
| 436.cactusADM: | -fast(cc) -fast(f90) -xipo=2 -fma=fused -xpagesize=4M -xprefetch=latx:2 -xalias_level=std -xprefetch_level=3 -xprefetch_auto_type=indirect_array_access -xautopar -xreduction |
| 454.calculix: | -fast(cc) -fast(f90) -xipo=2 -fma=fused -xalias_level=std |
| 481.wrf: | basepeak = yes |