White Papers

Abaqus Performance
16 Dell EMC Ready Solution for HPC Digital Manufacturing—Dassault Systѐmes’ Simulia Abaqus Performance
For all of the models test, substantial performance gains can be made using multiple domains per node,
where using 8 domains (6 threads per domain) delivers the optimal performance. Users are encouraged to
examine this option with their models to determine the optimal value. An even number is preferred, since it
would allow MPI processor binding to be enabled to further improve performance. There may be an increase
in the amount of memory required to minimize I/O when more than a single domain is placed on a node and
one needs to be careful to avoid “out-of-core” solutions, causing potentially significant I/O activity, decreasing
the overall performance. The user can examine the domain memory requirements to minimize I/O in the .dat
file to make sure this does not occur.
Figure 7 shows the performance on the larger standard and explicit benchmarks mentioned above on a 6252
based system, where the number of cores used was varied from 8 to 48 cores.
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
950
1000
1050
1100
1150
1200
S2A S4B S4D S6
Solver Elapsed Time (sec)
Figure 6: "mp_host_split" Performance
1 2 4 8
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
8-core 16-core 24-core 32-core 40-core 48-core
Performance (relative to 8
-core)
Figure 7: Single Server Parallel Performance
S3D S4B S4D S6 E1 E3 E6