SPEC® MPIM2007 Result

Copyright 2006-2010 Standard Performance Evaluation Corporation

Lenovo Global Technology

ThinkSystem SR665
(AMD EPYC 7763, 2.45 GHz)

MPI2007 license: 28 Test date: Mar-2021
Test sponsor: Lenovo Global Technology Hardware Availability: Mar-2021
Tested by: Lenovo Global Technology Software Availability: Mar-2021
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 256 28.9 54.1 29.1 53.7 29.7 52.8 256 28.9 54.1 29.1 53.7 29.7 52.8
107.leslie3d 256 115   45.3 115   45.2 116   45.2 256 115   45.3 115   45.2 116   45.2
113.GemsFDTD 256 191   33.1 190   33.1 191   33.0 256 191   33.1 190   33.1 191   33.0
115.fds4 256 31.9 61.2 31.8 61.3 31.6 61.7 256 31.9 61.2 31.8 61.3 31.6 61.7
121.pop2 256 191   21.7 190   21.8 189   21.8 256 191   21.7 190   21.8 189   21.8
122.tachyon 256 43.9 63.7 43.3 64.6 44.5 62.9 256 43.9 63.7 43.3 64.6 44.5 62.9
126.lammps 256 94.9 30.7 95.5 30.5 94.6 30.8 256 94.9 30.7 95.5 30.5 94.6 30.8
127.wrf2 256 75.4 103   75.5 103   75.4 103   256 75.4 103   75.5 103   75.4 103  
128.GAPgeofem 256 23.9 86.2 23.5 88.0 24.4 84.7 256 23.9 86.2 23.5 88.0 24.4 84.7
129.tera_tf 256 54.7 50.6 54.3 51.0 54.4 50.9 256 54.7 50.6 54.3 51.0 54.4 50.9
130.socorro 256 67.6 56.5 67.5 56.5 71.1 53.7 256 67.6 56.5 67.5 56.5 71.1 53.7
132.zeusmp2 256 57.2 54.3 57.9 53.6 56.9 54.6 256 57.2 54.3 57.9 53.6 56.9 54.6
137.lu 256 45.5 80.7 45.6 80.5 44.7 82.2 256 45.5 80.7 45.6 80.5 44.7 82.2
Hardware Summary
Type of System: Homogeneous
Compute Node: ThinkSystem SR665
Interconnect: Mellanox ConnectX-6 HDR
File Server Node: NFS
Total Compute Nodes: 2
Total Chips: 4
Total Cores: 256
Total Threads: 256
Total Memory: 2 TB
Base Ranks Run: 256
Minimum Peak Ranks: 256
Maximum Peak Ranks: 256
Software Summary
C Compiler: AMD Optimizing C Compiler for Linux
Version 3.0.0 Build 2020_12_10
C++ Compiler: AMD Optimizing C++ Compiler for Linux
Version 3.0.0 Build 2020_12_10
Fortran Compiler: AMD Optimizing Fortran Compiler for Linux
Version 3.0.0 Build 2020_12_10
Base Pointers: 64-bit
Peak Pointers: Not Applicable
MPI Library: Open MPI Library
Version 4.1.0
Other MPI Info: None
Pre-processors: No
Other Software: None

Node Description: ThinkSystem SR665

Hardware
Number of nodes: 2
Uses of the node: compute
Vendor: Lenovo Global Technology
Model: SR665
CPU Name: AMD EPYC 7763
CPU(s) orderable: 1-2 chips
Chips enabled: 2
Cores enabled: 128
Cores per chip: 64
Threads per core: 1
CPU Characteristics: None
CPU MHz: 2450
Primary Cache: 32 KB I + 32 KB D on chip per core
Secondary Cache: 512 KB I+D on chip per core
L3 Cache: 256 MB I+D on chip per chip
32 MB shared / 8 cores
Other Cache: None
Memory: 1 TB (16 x 64 GB 2Rx4 PC4-3200AA-R)
Disk Subsystem: 1 x 480 GB SATA 2.5" SSD
Other Hardware: None
Adapter: Mellanox ConnectX-6 HDR Infiniband
Number of Adapters: 1
Slot Type: PCI-Express 4.0 x16
Data Rate: 200 Gbs/s
Ports Used: 1
Interconnect Type: Mellanox ConnectX-6 HDR Infiniband Adapter
Software
Adapter: Mellanox ConnectX-6 HDR Infiniband
Adapter Driver: 5.2-1.0.4
Adapter Firmware: 20.25.2006
Operating System: Red Hat Enterprise Linux Server release 8.3
4.18.0-240.el8.x86_64
Local File System: xfs
Shared File System: None
System State: Multi-user, run level 3
Other Software: None

Node Description: NFS

Hardware
Number of nodes: 1
Uses of the node: Fileserver
Vendor: Lenovo Global Technology
Model: ThinkSystem SR665
CPU Name: AMD EPYC 7763 CPU
CPU(s) orderable: 1-2 chips
Chips enabled: 2
Cores enabled: 128
Cores per chip: 64
Threads per core: 1
CPU Characteristics: None
CPU MHz: 2450
Primary Cache: 32 KB I + 32 KB D on chip per core
Secondary Cache: 512 KB I+D on chip per core
L3 Cache: 256 MB I+D on chip per chip
32 MB shared / 8 cores
Other Cache: None
Memory: 1 TB (16 x 64 GB 2Rx4 PC4-3200AA-R)
Disk Subsystem: 1 x 480 GB SATA 2.5" SSD
Other Hardware: None
Adapter: Mellanox ConnectX-6 HDR Infiniband
Number of Adapters: 1
Slot Type: PCI-Express 4.0 x16
Data Rate: 200 Gb/s
Ports Used: 1
Interconnect Type: Mellanox ConnectX-6 HDR Infiniband
Software
Adapter: Mellanox ConnectX-6 HDR Infiniband
Adapter Driver: 5.2-1.0.4
Adapter Firmware: 20.25.2006
Operating System: Red Hat Enterprise Linux Server release 8.3
Local File System: None
Shared File System: NFS
System State: Multi-User, run level 3
Other Software: None

Interconnect Description: Mellanox ConnectX-6 HDR

Hardware
Vendor: Mellanox
Model: Infiniband HDR 200Gb/s Switch
Switch Model: QM8700 Series
Number of Switches: 1
Number of Ports: 40
Data Rate: 200 Gb/s
Firmware: 3.9.0606
Topology: Mesh
Primary Use: MPI Traffic

Submit Notes

The config file option 'submit' was used.

General Notes

 MPI startup command:
   mpiexec command was used to start MPI jobs.
 RAM configuration:
   Compute nodes have 1 x 64 GB RDIMM on each memory channel.
 Add "idle=poll" into grub
 BIOS settings:
   Operating Mode : Maximum Performance Mode
   Hyper-Threading Technology (SMT): Enabled
   NPS4
 Yes: The test sponsor attests, as of date of publication,
 that CVE-2017-5754 (Meltdown) is mitigated in the system as tested and documented.
 Yes: The test sponsor attests, as of date of publication,
 that CVE-2017-5753 (Spectre variant 1) is mitigated in the system as tested and documented.
 Yes: The test sponsor attests, as of date of publication,
 that CVE-2017-5715 (Spectre variant 2) is mitigated in the system as tested and documented.

Base Compiler Invocation

C benchmarks:

 mpicc 

C++ benchmarks:

126.lammps:  mpicxx 

Fortran benchmarks:

 mpifort 

Benchmarks using both Fortran and C:

 mpicc   mpifort 

Base Portability Flags

121.pop2:  -DSPEC_MPI_CASE_FLAG 
126.lammps:  -DMPICH_IGNORE_CXX_SEEK 
127.wrf2:  -DSPEC_MPI_CASE_FLAG   -DSPEC_MPI_LINUX   -Wno-return-type 

Base Optimization Flags

C benchmarks:

 -Ofast   -flto   -ffast-math   -march=znver3   -mavx2   -lamdlibm 

C++ benchmarks:

126.lammps:  -Ofast   -flto   -ffast-math   -march=znver3   -mavx2   -lamdlibm 

Fortran benchmarks:

 -Ofast   -flto   -ffast-math   -march=znver3   -mavx2   -funroll-loops   -lamdlibm 

Benchmarks using both Fortran and C:

 -Ofast   -flto   -ffast-math   -march=znver3   -mavx2   -lamdlibm   -funroll-loops 

Peak Optimization Flags

C benchmarks:

104.milc:  basepeak = yes 
122.tachyon:  basepeak = yes 

C++ benchmarks:

126.lammps:  basepeak = yes 

Fortran benchmarks:

107.leslie3d:  basepeak = yes 
113.GemsFDTD:  basepeak = yes 
129.tera_tf:  basepeak = yes 
137.lu:  basepeak = yes 

Benchmarks using both Fortran and C:

115.fds4:  basepeak = yes 
121.pop2:  basepeak = yes 
127.wrf2:  basepeak = yes 
128.GAPgeofem:  basepeak = yes 
130.socorro:  basepeak = yes 
132.zeusmp2:  basepeak = yes 

The flags file that was used to format this result can be browsed at
http://www.spec.org/mpi2007/flags/AMD_flags.20210315.html.

You can also download the XML flags source by saving the following link:
http://www.spec.org/mpi2007/flags/AMD_flags.20210315.xml.