Skip to content
Snippets Groups Projects
mpi.md 5.78 KiB
Newer Older
  • Learn to ignore specific revisions
  • Lukáš Krupčík's avatar
    Lukáš Krupčík committed
    # MPI
    
    ## Setting Up MPI Environment
    
    The Salomon cluster provides several implementations of the MPI library:
    
    | MPI Library       | Thread support                                                                                         |
    | ----------------- | ------------------------------------------------------------------------------------------------------ |
    | **Intel MPI 4.1** | Full thread support up to                                        `MPI_THREAD_MULTIPLE`                  |
    | **Intel MPI 5.0** | Full thread support up to                                        `MPI_THREAD_MULTIPLE`                  |
    | OpenMPI 1.8.6     | Full thread support up to                                       `MPI_THREAD_MULTIPLE`, MPI-3.0 support |
    | SGI MPT 2.12      |                                                                                                        |
    
    MPI libraries are activated via the environment modules.
    
    Look up the modulefiles/mpi section in `ml av`:
    
    ```console
    $ ml av
        ------------------------------ /apps/modules/mpi -------------------------------
        impi/4.1.1.036-iccifort-2013.5.192
        impi/4.1.1.036-iccifort-2013.5.192-GCC-4.8.3
        impi/5.0.3.048-iccifort-2015.3.187
        impi/5.0.3.048-iccifort-2015.3.187-GNU-5.1.0-2.25
        MPT/2.12
        OpenMPI/1.8.6-GNU-5.1.0-2.25
    ```
    
    There are default compilers associated with any particular MPI implementation. The defaults may be changed; the MPI libraries may be used in conjunction with any compiler. The defaults are selected via the modules in the following way:
    
    | Module                                   | MPI        | Compiler suite |
    | ---------------------------------------- | ---------- | -------------- |
    | impi-5.0.3.048-iccifort- Intel MPI 5.0.3 | 2015.3.187 |                |
    | OpenMP-1.8.6-GNU-5.1.0-2 OpenMPI 1.8.6   | .25        |                |
    
    Examples:
    
    ```console
    $ ml gompi/2015b
    ```
    
    In this example, we activate the latest OpenMPI with the latest GNU compilers (OpenMPI 1.8.6 and GCC 5.1). For more information about toolchains, see the [Environment and Modules][1] section.
    
    To use OpenMPI with the Intel compiler suite, use:
    
    ```console
    $ ml iompi/2015.03
    ```
    
    In this example, the OpenMPI 1.8.6 using the Intel compilers is activated. It uses the `iompi` toolchain.
    
    ## Compiling MPI Programs
    
    After setting up your MPI environment, compile your program using one of the MPI wrappers:
    
    ```console
    $ mpicc -v
    $ mpif77 -v
    $ mpif90 -v
    ```
    
    When using Intel MPI, use the following MPI wrappers:
    
    ```console
    $ mpicc
    $ mpiifort
    ```
    
    Wrappers `mpif90` and `mpif77` provided by Intel MPI are designed for GCC and GFortran. You might be able to compile MPI code by them even with Intel compilers, but you might run into problems (for example, native MIC compilation with `-mmic` does not work with `mpif90`).
    
    Example program:
    
    ```cpp
    // helloworld_mpi.c
    #include <stdio.h>
    
    #include<mpi.h>
    
    int main(int argc, char **argv) {
    
    int len;
    int rank, size;
    char node[MPI_MAX_PROCESSOR_NAME];
    
    // Initiate MPI
    MPI_Init(&argc, &argv);
    MPI_Comm_rank(MPI_COMM_WORLD,&rank);
    MPI_Comm_size(MPI_COMM_WORLD,&size);
    
    // Get hostame and print
    MPI_Get_processor_name(node,&len);
    printf("Hello world! from rank %d of %d on host %sn",rank,size,node);
    
    // Finalize and exit
    MPI_Finalize();
    
    return 0;
    }
    ```
    
    Compile the above example with:
    
    ```console
    $ mpicc helloworld_mpi.c -o helloworld_mpi.x
    ```
    
    ## Running MPI Programs
    
    The MPI program executable must be compatible with the loaded MPI module.
    Always compile and execute using the very same MPI module.
    
    It is strongly discouraged to mix MPI implementations. Linking an application with one MPI implementation and running `mpirun`/`mpiexec` from another implementation may result in unexpected errors.
    
    The MPI program executable must be available within the same path on all nodes. This is automatically fulfilled on the /home and /scratch filesystem. You need to preload the executable if running on the local scratch /lscratch filesystem.
    
    ### Ways to Run MPI Programs
    
    The optimal way to run an MPI program depends on its memory requirements, memory access pattern and communication pattern.
    
    !!! note
        Consider these ways to run an MPI program:
        1. One MPI process per node, 24 threads per process
        2. Two MPI processes per node, 12 threads per process
        3. 24 MPI processes per node, 1 thread per process.
    
    **One MPI** process per node, using 24 threads, is most useful for memory demanding applications that make good use of processor cache memory and are not memory-bound. This is also a preferred way for communication intensive applications as one process per node enjoys full bandwidth access to the network interface.
    
    **Two MPI** processes per node, using 12 threads each, bound to processor socket is most useful for memory bandwidth-bound applications such as BLAS1 or FFT with scalable memory demand. However, note that the two processes will share access to the network interface. The 12 threads and socket binding should ensure maximum memory access bandwidth and minimize communication, migration, and NUMA effect overheads.
    
    !!! note
        Important! Bind every OpenMP thread to a core!
    
    In the previous two cases with one or two MPI processes per node, the operating system might still migrate OpenMP threads between cores. You want to avoid this by setting the `KMP_AFFINITY` or `GOMP_CPU_AFFINITY` environment variables.
    
    **24 MPI** processes per node, using 1 thread each bound to a processor core is most suitable for highly scalable applications with low communication demand.
    
    ### Running OpenMPI
    
    The [OpenMPI 1.8.6][a] is based on OpenMPI. Read more on [how to run OpenMPI][2].
    
    The Intel MPI may run on the [Intel Xeon Ph][3] accelerators as well. Read more on [how to run Intel MPI on accelerators][3].
    
    [1]: ../../modules-matrix.md
    [2]: running_openmpi.md
    [3]: ../intel/intel-xeon-phi-salomon.md
    
    [a]: http://www.open-mpi.org/