SPEC® MPIM2007 Result

Copyright 2006-2010 Standard Performance Evaluation Corporation

SGI

SGI Altix 4700 Bandwidth System (Itanium 2
Processor 9040 1.6GHz/18M)

SPECmpiM_peak2007 = Not Run

MPI2007 license: 4 Test date: May-2007
Test sponsor: SGI Hardware Availability: Jul-2006
Tested by: SGI Software Availability: Apr-2007
Benchmark results graph

Results Table

Benchmark Base Peak
Ranks Seconds Ratio Seconds Ratio Seconds Ratio Ranks Seconds Ratio Seconds Ratio Seconds Ratio
Results appear in the order in which they were run. Bold underlined text indicates a median measurement.
104.milc 64 308 5.07 308 5.08
107.leslie3d 64 436 12.0  433 12.0 
113.GemsFDTD 64 632 9.99 631 9.99
115.fds4 64 383 5.09 392 4.98
121.pop2 64 756 5.46 755 5.47
122.tachyon 64 974 2.87 974 2.87
126.lammps 64 815 3.58 814 3.58
127.wrf2 64 696 11.2  694 11.2 
128.GAPgeofem 64 314 6.58 313 6.60
129.tera_tf 64 582 4.75 582 4.75
130.socorro 64 256 14.9  258 14.8 
132.zeusmp2 64 550 5.64 548 5.66
137.lu 64 257 14.3  256 14.4 
Hardware Summary
Type of System: SMP
Compute Node: SMP
File Server Node: SMP
Total Compute Nodes: 1
Total Chips: 32
Total Cores: 64
Total Threads: 64
Total Memory: 256 GB
Base Ranks Run: 64
Minimum Peak Ranks: --
Maximum Peak Ranks: --
Software Summary
C Compiler: Intel C Itanium Compiler for Itanium-based
Applications Version 9.1 (Build 20070320)
C++ Compiler: Intel C++ Itanium Compiler for Itanium-based
Applications Version 9.1 (Build 20070320)
Fortran Compiler: Intel Fortran Itanium Compiler for Itanium-based
Applications Version 9.1 (Build 20070320)
Base Pointers: 64-bit
Peak Pointers: 64-bit
MPI Library: SGI Message Passing Toolkit (MPT) Version 1.15
Other MPI Info: None
Pre-processors: None
Other Software: None

Node Description: SMP

Hardware
Number of nodes: 1
Uses of the node: compute, file server
Vendor: SGI
Model: SGI Altix 4700 Bandwidth System (Itanium 2
Processor 9040 1.6GHz/18M)
CPU Name: Dual-Core Intel Itanium 2 9040
CPU(s) orderable: 1-512 chips
Chips enabled: 32
Cores enabled: 64
Cores per chip: 2
Threads per core: 1
CPU Characteristics: 533MHz FSB
CPU MHz: 1600
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: 9 MB I+D on chip per core
Other Cache: None
Memory: 256 GB (8*1GB DDR2-400 DIMMS per 2 core module)
Disk Subsystem: 36 x 73 GB FibreChannel (Seagate Cheetah 15k rpm)
Other Hardware: None
Adapter: None
Number of Adapters: 0
Slot Type: Not applicable
Data Rate: Not applicable
Ports Used: 0
Interconnect Type: None
Software
Adapter: None
Adapter Driver: Not applicable
Adapter Firmware: Not applicable
Operating System: SUSE Linux Enterprise Server 10 +
SGI ProPack 5 Service Pack 1
Local File System: 36 x 73 GB FibreChannel (Seagate Cheetah 15k rpm)
Shared File System: None
System State: Multi-user
Other Software: None

General Notes

 setenv MPI_DSM_DISTRIBUTE 1
      Ensures that each MPI process gets a unique CPU and physical
      memory on the node with which that CPU is associated.  The
      CPUs are chosen by simply starting at cpuset-relative CPU 0
      and incrementing until all MPI processes have been forked.
 setenv MPI_REQUEST_MAX 65536
      Determines the maximum number of nonblocking sends and
      receives that can simultaneously exist for any single MPI
      process.  MPI generates an error message if this limit
      (or the default, if not set) is exceeded.  Default:  16384
 limit stacksize unlimited
      Removes limits on the maximum size of the automatically-
      extended stack region of the current process and each
      process it creates.

Base Compiler Invocation

C benchmarks:

 icc 

C++ benchmarks:

126.lammps:  icpc 

Fortran benchmarks:

 ifort 

Benchmarks using both Fortran and C:

 icc   ifort 

Base Portability Flags

121.pop2:  -DSPEC_MPI_CASE_FLAG 
127.wrf2:  -DSPEC_MPI_LINUX   -DSPEC_MPI_CASE_FLAG 

Base Optimization Flags

C benchmarks:

 -O3   -ipo   -IPF-fp-relaxed   -lmpi 

C++ benchmarks:

126.lammps:  -O3   -ipo   -IPF-fp-relaxed   -ansi-alias   -lmpi 

Fortran benchmarks:

 -O3   -ipo   -IPF-fp-relaxed   -lmpi 

Benchmarks using both Fortran and C:

 -O3   -ipo   -IPF-fp-relaxed   -lmpi 

The flags file that was used to format this result can be browsed at
http://www.spec.org/mpi2007/flags/Intel-ic91-ipf.html.

You can also download the XML flags source by saving the following link:
http://www.spec.org/mpi2007/flags/Intel-ic91-ipf.xml.