For large models, you can use the shared memory parallel processing (SMP) or the massively parallel processing (MPP) capabilities of ANSYS LS-DYNA to shorten the elapsed time necessary to run an analysis. To use either of these features, you must purchase the appropriate number of ANSYS LS-DYNA Parallel licenses. The Parallel license incorporates both SMP and MPP capabilities. Please contact your ANSYS sales representative for more information on purchasing the appropriate licenses.
The shared memory parallel processing capabilities allow you to distribute model-solving power over multiple processors on the same machine. To use this feature you must have a machine with at least two processors, and you must have an ANSYS LS-DYNA Parallel license for each processor beyond the first one.
To use the SMP capabilities you need to issue the EDRUN command
to specify an SMP run as well as the number of processors to use
(EDRUN,SMP,,Ncpu).
When you are using shared memory parallel processing, the calculations may be executed in different order, depending on CPU availability and the workload on each CPU. Because of this, you may see slight differences in the results when you run the same job multiple times. To avoid these differences, you can specify that consistency be maintained (EDRUN,SMP,1). Maintaining consistency can result in an increase of up to 15% in CPU time.
If you are using all CPUs on a platform (ANSYS, Inc. recommends always running one fewer CPU than is available), you should close all other applications on those CPUs. Any calculations or machine resources that compete with the ANSYS LS-DYNA application running in SMP mode will reduce the SMP performance significantly.
The SMP capability is expected to scale linearly when used with up to 5 processors. When used with more than 5 processors, any additional scalability is minimal; the larger the problem, the higher the potential for scalability.) If you are using more than 5 processors, we recommend using the MPP capability, which offers excellent scalability regardless of the number of processors.
The massively parallel processing (MPP) capabilities of ANSYS LS-DYNA allow you to run the LS-DYNA solver over a cluster of machines or use multiple processors on a single machine. To use the LS-DYNA MPP feature, you must have an ANSYS LS-DYNA Parallel license for each processor beyond the first one.
Before running an analysis using LS-DYNA MPP, you must configure your system properly. You must have supported MPI software correctly installed, and the machines running LS-DYNA MPP must be properly configured. The configuration procedures for LS-DYNA MPP are the same as for Distributed ANSYS. For more information, see Configuring Distributed ANSYS in the Parallel Processing Guide.
Not all platforms and MPI software listed in Prerequisites for Running Distributed ANSYS are supported by LS-DYNA MPP. For details, refer to the discussion of LS-DYNA MPP MPI support found in that section.
To run LS-DYNA MPP you can use the High Performance Computing Tab in the ANSYS Launcher or run the LS-DYNA executable via the command line using the correct arguments (see Activating Distributed ANSYS in the Parallel Processing Guide for more information). Running LS-DYNA MPP from the command line is similar to running Distributed ANSYS via the command line. The arguments that are common to both programs are:
| -dis | Enables distributed processing (MPP) version of LS-DYNA |
| -np | Specifies the number of processors to use when running LS-DYNA MPP |
| -machines [1] | Specifies the machines on which to run an LSDYNA MPP analysis |
| -mpifile | Specifies an existing MPI file |
The -dis argument is required in order to run LS-DYNA MPP. Choose one of the three remaining arguments (-np, -machines, or -mpifile) to indicate how the analysis will be distributed over multiple processors or machines. Note that the -np and -machines arguments are mutually exclusive; -machines will override -np. See Starting Distributed ANSYS via Command Line in the Parallel Processing Guide for information on how to use these command line options.
Other command line options (such as I/O and other LS-DYNA specific arguments) are unique to LS-DYNA. Because you are using the command line, you must follow the procedure outlined in Editing the LS-DYNA Input File. This involves using the EDWRITE command to create a Jobname.K input file, then using the lsdyna182 script with the appropriate arguments.
Note: The procedure described in Editing the LS-DYNA Input File indicates that you should go back to the ANSYS LS-DYNA program to view the analysis results. This step does not apply to an LS-DYNA MPP analysis.
The following is an example of a command line used to run LS-DYNA MPP on a Linux system. This example assumes you have ANSYS LS-DYNA at ANSYS Release 18.2, your input file is "file.k", and you are using two machines (one with four processors and one with two processors).
lsdyna182 -dis -machines machine1:4:machine2:2 i=file.k pr=dysmp
The output files produced by LS-DYNA MPP are standard LS-DYNA output files and are not compatible with ANSYS. Therefore, you must use LS-DYNA postprocessors to view the results.