Page History

...

Beyond the different compiler flavours in offer, we can also choose different MPI implementations for our MPI parallel programs. On the Atos HPCF and ECS, we can choose from the following implementations:

Implementation	Module	Description
OpenMPI	`openmpi`	Standard OpenMPI implementation provided by Atos
Intel MPI	`intel-mpi`	Intel's MPI implementation based on MPICH. Part of part of the Intel OneAPI distribution
HPC-X OpenMPI	`hpcx-openmpi`	NVIDIA's optimised flavour of OpenMPI. This is the recommended option

For the next exercise, we will use this adapted hello world code for MPI.

...

Reset your environment with:
No Format
module reset

With your favourite editor, create the file mpiversions.c with the code above, and compile it into the executable mpiversions. Hint: You may use the module hpcx-openmpi.

Expand

title	Solution

In order to compile MPI parallel programs, we need to use the MPI wrappers such as mpicc for C, mpicxx / mpic++ for C++ or mpif77 / mpif90 / mpifort for Fortran code. They will in turn call the corresponding compiler loaded in your environment, and will make sure all the necessary flags to compile and link agains the MPI library are set.

These wrappers are made available when you load one of the MPI modules in the system. We will use the default hpcx-openmpi.

Since we are building a C program, we will use mpicc:

No Format
module load hpcx-openmpi mpicc -o mpiversions mpiversions.c

Write a small batch job that will compile and run the program using 2 processors and submit it to the batch system:

Expand

title	Solution

You may write a job script similar to the following

Code Block

language	bash
title	mpiversions.sh

#!/bin/bash
#SBATCH -J mpiversions
#SBATCH -o %x.out
#SBATCH -n 2

module load hpcx-openmpi
mpicc -o mpiversions mpiversions.c
srun ./mpiversions

You may then submit it to the batch system with sbatch:

No Format
sbatch mpiversions.sh

Inspect the output to check what versions of Compiler and MPI are reported

Tweak the previous job to build and run the mpiversions program with as many combinations of compiler families and MPI implementations as you can.

Expand

title	Solution

You may amend your existing mpiversions.sh script with a couple of loops on the prgenvs and MPI implementations:

Code Block

language	bash
title	mpiversions.sh

#!/bin/bash
#SBATCH -J mpiversions
#SBATCH -o %x.out
#SBATCH -n 2

for pe in gnu intel intel-llvm amd; do
    for mpi in hpcx-openmpi intel-mpi openmpi; do
        module load prgenv/$pe $mpi
        module list 
        mpicc -o mpiversions mpiversions.c
        srun ./mpiversions
        echo "******************"
    done
done

You may then submit it to the batch system with sbatch:

No Format
sbatch mpiversions.sh

Inspect again the output to check what versions of Compiler and MPI are reported.

...

Space shortcuts

Page tree

Versions Compared

Old Version 8

New Version 9

Key