SPEC® MPIM2007 Result

Copyright 2006-2010 Standard Performance Evaluation Corporation

Colfax International

Barcelona Cluster

SPECmpiM_peak2007 = Not Run

MPI2007 license: 021 Test date: Sep-2007
Test sponsor: Scali, Inc. Hardware Availability: Sep-2007
Tested by: Scali, Inc. Software Availability: Aug-2007

SPEC has determined that this result was not in compliance
with the SPEC MPI2007 run and reporting rules. Specifically,
the processor vendor reported that the processor would not
meet the SPEC HPG requirements for continued availability.

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 NA NA NA NA NA NA NA
107.leslie3d NA NA NA NA NA NA NA
113.GemsFDTD NA NA NA NA NA NA NA
115.fds4 NA NA NA NA NA NA NA
121.pop2 NA NA NA NA NA NA NA
122.tachyon NA NA NA NA NA NA NA
126.lammps NA NA NA NA NA NA NA
127.wrf2 NA NA NA NA NA NA NA
128.GAPgeofem NA NA NA NA NA NA NA
129.tera_tf NA NA NA NA NA NA NA
130.socorro NA NA NA NA NA NA NA
132.zeusmp2 NA NA NA NA NA NA NA
137.lu NA NA NA NA NA NA NA
Hardware Summary
Type of System: Homogenous
Compute Node: H8DMU
Interconnects: mpiComm
GBEthernet
File Server Node: X4100
Total Compute Nodes: 8
Total Chips: 16
Total Cores: 64
Total Threads: 64
Total Memory: 128 GB
Base Ranks Run: 64
Minimum Peak Ranks: --
Maximum Peak Ranks: --
Software Summary
C Compiler: QLogic PathScale C Compiler 3.0
C++ Compiler: QLogic PathScale C++ Compiler 3.0
Fortran Compiler: QLogic PathScale Fortran Compiler 3.0
Base Pointers: 64-bit
Peak Pointers: Not Applicable
MPI Library: Scali MPI Connect 5.5
Other MPI Info: OFED 1.2.5 ibverbs
Pre-processors: No
Other Software: None

Node Description: H8DMU

Hardware
Number of nodes: 8
Uses of the node: compute
Vendor: SuperMicro
Model: H8DMU+
CPU Name: AMD Opteron CPU 2350
CPU(s) orderable: 1 or 2 chips
Chips enabled: 2
Cores enabled: 8
Cores per chip: 4
Threads per core: 1
CPU Characteristics: Quad-Core AMD Opteron Processor 2350 (Barcelona)
CPU MHz: 2000
Primary Cache: 64 KB I + 64 KB D on chip per core
Secondary Cache: 512 KB I+D on chip per core
L3 Cache: 2 MB I+D on chip per chip
Other Cache: None
Memory: 16 GB (8 x 2 GB DDR2 667 Micron)
Disk Subsystem: 400GB Seagate SATA, 7200RPM
Other Hardware: None
Adapter: Ethernet controller: nVidia Corporation MCP55
Ethernet
Number of Adapters: 2
Slot Type: PCIe x8
Data Rate: 1 Gbps Ethernet
Ports Used: 1
Interconnect Type: Gigabit Ethernet
Adapter: Mellanox ConnectX DDR, board id MT_04A0110002
Number of Adapters: 1
Slot Type: PCIe x8
Data Rate: InfiniBand 4x DDR
Ports Used: 1
Interconnect Type: Infiniband
Software
Adapter: Ethernet controller: nVidia Corporation MCP55
Ethernet
Adapter Driver: OS default
Adapter Firmware: Unknown
Adapter: Mellanox ConnectX DDR, board id MT_04A0110002
Adapter Driver: OFED 1.2.5
Adapter Firmware: 2.2.000
Operating System: CentOS release 4.5 (Final), 2.6.9-55.0.2.ELsmp
Local File System: ext3
Shared File System: NFS
System State: Multi-user
Other Software: None

Node Description: X4100

Hardware
Number of nodes: 1
Uses of the node: fileserver
Vendor: Sun Microsystems, Inc.
Model: Sun Fire X4100
CPU Name: AMD Opteron 285
CPU(s) orderable: 1-2 chip
Chips enabled: 2
Cores enabled: 4
Cores per chip: 2
Threads per core: 1
CPU Characteristics: Dual Core AMD Opteron Processor 285
CPU MHz: 2600
Primary Cache: 64 KB I + 64 KB D on chip per core
Secondary Cache: 1 MB I+D on chip per chip
L3 Cache: None
Other Cache: None
Memory: 8 GB (8 x 1GB DDR2/667 ECC registered DIMMs)
Disk Subsystem: 2x SAS 10k RPM mirrored
Other Hardware: None
Adapter: Intel Corporation 82546EB Gigabit Ethernet
Controller
Number of Adapters: 4
Slot Type: PCIe x8
Data Rate: 1 Gbps Ethernet
Ports Used: 1
Interconnect Type: Gigabit Ethernet
Software
Adapter: Intel Corporation 82546EB Gigabit Ethernet
Controller
Adapter Driver: OS default
Adapter Firmware: Unknown
Operating System: CentOS release 4.5 (Final), 2.6.9-55.0.2.ELsmp
Local File System: ext3
Shared File System: NFS
System State: Multi-user
Other Software: None

Interconnect Description: mpiComm

Hardware
Vendor: Voltaire
Model: Voltaire 9024D 24 ports DDR switch
Switch Model: 9024D
Number of Switches: 1
Number of Ports: 24
Data Rate: InfiniBand 4x DDR
Firmware: Unknown
Topology: single switch (star)
Primary Use: MPI traffic

Interconnect Description: GBEthernet

Hardware
Vendor: Nortel
Model: Nortel Networks Baystack 5510 Gigabit Ethernet
switch
Switch Model: 5510
Number of Switches: 1
Number of Ports: 24
Data Rate: 1 Gbps Ethernet
Firmware: fw: 1.0.0.16, sw: v3.0.1.00
Topology: Single Switch
Primary Use: file system traffic

Submit Notes

Scali MPI Connect's mpirun wrapper has been
used to submit the jobs. Description of switches:
  -npn 8: launch 8 processes per node.
  -rsh rsh: use rsh as method to connect to nodes.
  -mstdin none: do not connect the processes' STDIN
     to anything.
  -q: quiet mode, no output from launcher.
  -machinefile: file selecting the hosts to run on.

General Notes

Scali, Inc has executed the benchmark on AMD
Development Center. We are grateful for the support from
AMD and in particular Joshua Mora and Brian Taylor in order to
finalize the submissions.

Base Compiler Invocation

C benchmarks:

 /opt/scali/bin/mpicc -ccl pathcc 

C++ benchmarks:

126.lammps:  /opt/scali/bin/mpicc -ccl pathCC 

Fortran benchmarks:

 /opt/scali/bin/mpif77 -ccl pathf90 

Benchmarks using both Fortran and C:

 /opt/scali/bin/mpicc -ccl pathcc   /opt/scali/bin/mpif77 -ccl pathf90 

Base Portability Flags

104.milc:  -DSPEC_MPI_LP64 
115.fds4:  -DSPEC_MPI_LC_TRAILING_DOUBLE_UNDERSCORE   -DSPEC_MPI_LP64 
121.pop2:  -DSPEC_MPI_DOUBLE_UNDERSCORE   -DSPEC_MPI_LP64 
122.tachyon:  -DSPEC_MPI_LP64 
127.wrf2:  -DF2CSTYLE   -DSPEC_MPI_DOUBLE_UNDERSCORE   -DSPEC_MPI_LINUX   -DSPEC_MPI_LP64 
128.GAPgeofem:  -DSPEC_MPI_LP64 
130.socorro:  -fno-second-underscore   -DSPEC_MPI_LP64 
132.zeusmp2:  -DSPEC_MPI_LP64 

Base Optimization Flags

C benchmarks:

 -march=core   -Ofast   -OPT:malloc_alg=1 

C++ benchmarks:

126.lammps:  -march=core   -O3   -OPT:Ofast   -CG:local_fwd_sched=on 

Fortran benchmarks:

 -march=core   -O3   -OPT:Ofast   -OPT:malloc_alg=1   -LANG:copyinout=off 

Benchmarks using both Fortran and C:

 -march=core   -Ofast   -OPT:malloc_alg=1   -O3   -OPT:Ofast   -LANG:copyinout=off 

Base Other Flags

C benchmarks:

 -IPA:max_jobs=4 

C++ benchmarks:

126.lammps:  -IPA:max_jobs=4 

Fortran benchmarks:

 -IPA:max_jobs=4 

Benchmarks using both Fortran and C:

 -IPA:max_jobs=4 

The flags file that was used to format this result can be browsed at
http://www.spec.org/mpi2007/flags/MPI2007_flags.20071107.html.

You can also download the XML flags source by saving the following link:
http://www.spec.org/mpi2007/flags/MPI2007_flags.20071107.xml.