OpenMP - Revision history

Shtsai: /* Compiling OpenMP code on Sapelo2 */

2025-07-31T17:16:24Z

Compiling OpenMP code on Sapelo2

← Older revision		Revision as of 13:16, 31 July 2025
Line 13:		Line 13:

	\|-		\|-
	\| ~~PGI~~ \|\| ~~pgcc~~, ~~pgCC, pgfortran~~, etc. \|\| -mp		\| NVHPC \|\| nvc, nvfortran, etc. \|\| -mp
	\|-		\|-
	\| Intel \|\| icc, icpc, ifort \|\| -openmp		\| Intel \|\| icc, icpc, ifort \|\| -openmp
Line 22:		Line 22:

	For more information about the compilers on Sapelo2, please see [[Code Compilation on Sapelo2]].		For more information about the compilers on Sapelo2, please see [[Code Compilation on Sapelo2]].

	===Defining number of OpenMP threads===		===Defining number of OpenMP threads===

Moses at 15:48, 6 September 2023

2023-09-06T15:48:34Z

← Older revision		Revision as of 11:48, 6 September 2023
Line 51:		Line 51:

	<pre class="gcommand">		<pre class="gcommand">
	setenv OMP_PROC_BIND true		setenv OMP_PROC_BIND true
	</pre>		</pre>

Line 78:		Line 78:
	export OMP_PROC_BIND=true		export OMP_PROC_BIND=true

	module load foss/~~2019b~~ # load the appropriate module ~~file~~, e.g. foss/~~2019b~~		module load foss/2022a # load the appropriate module, e.g. foss/2022a

	time ./a.out		time ./a.out

Moses at 14:05, 8 November 2021

2021-11-08T14:05:54Z

← Older revision		Revision as of 10:05, 8 November 2021
Line 15:		Line 15:
	\| PGI \|\| pgcc, pgCC, pgfortran, etc. \|\| -mp		\| PGI \|\| pgcc, pgCC, pgfortran, etc. \|\| -mp
	\|-		\|-
	\| Intel \|\| icc, ifort \|\| -openmp		\| Intel \|\| icc, icpc, ifort \|\| -openmp
	\|-		\|-
	\| GNU \|\| gcc, g++, gfortran, etc. \|\| -fopenmp		\| GNU \|\| gcc, g++, gfortran, etc. \|\| -fopenmp
Line 39:		Line 39:
	</pre>		</pre>

			=== Setting thread-core binding ===
	===Setting thread-core binding===

	In general, an OpenMP code will run a lot more efficiently when threads are pinned to a given core on a compute node. Please set the environment variable '''OMP_PROC_BIND=true''' to specify threads' affinity policy (thread-core bind).		In general, an OpenMP code will run a lot more efficiently when threads are pinned to a given core on a compute node. Please set the environment variable '''OMP_PROC_BIND=true''' to specify threads' affinity policy (thread-core bind).

Moses: /* Defining number of OpenMP threads */

2021-11-08T13:57:30Z

Defining number of OpenMP threads

← Older revision		Revision as of 09:57, 8 November 2021
Line 25:		Line 25:
	===Defining number of OpenMP threads===		===Defining number of OpenMP threads===

	The number of OpenMP threads can be specified using the environment variable OMP_NUM_THREADS. For example, to define 4 threads:		The number of OpenMP threads can be specified using the environment variable '''OMP_NUM_THREADS'''. For example, to define 4 threads:

	For bash/sh:		For bash/sh:
Line 42:		Line 42:
	===Setting thread-core binding===		===Setting thread-core binding===

	In general, an OpenMP code will run a lot more efficiently when threads are pinned to a given core on a compute node. Please set the environment variable		In general, an OpenMP code will run a lot more efficiently when threads are pinned to a given core on a compute node. Please set the environment variable '''OMP_PROC_BIND=true''' to specify threads' affinity policy (thread-core bind).
	OMP_PROC_BIND=true to specify threads' affinity policy (thread-core bind).

	For bash/sh:		For bash/sh:
Line 81:		Line 80:
	export OMP_PROC_BIND=true		export OMP_PROC_BIND=true

	module load foss/2019b # load the appropriate module file, e.g. foss/2019b		module load foss/2019b # load the appropriate module file, e.g. foss/2019b

	time ./a.out		time ./a.out
	</pre>		</pre>

			'''Please note''' that the number of cores requested by this job ('''--cpus-per-task=4''') is the same as the number of OpenMP threads that the application will use ('''export OMP_NUM_THREADS=4''').
	'''Please note''' that the number of cores requested by this job (--cpus-per-task=4) is the same as the number of OpenMP threads that the application will use (export OMP_NUM_THREADS=4).

	===Sample job submission command on Sapelo2===		===Sample job submission command on Sapelo2===

Moses at 13:55, 8 November 2021

2021-11-08T13:55:34Z

← Older revision		Revision as of 09:55, 8 November 2021
Line 9:		Line 9:
	! scope="col" \| Compiler		! scope="col" \| Compiler
	! scope="col" \| Commands		! scope="col" \| Commands
	! scope="col" \| Option		! scope="col" \| Compile Option
	\|-		\|-

	\|-		\|-
	\| PGI \|\| pgcc, pgCC, ~~pgf90~~, etc. \|\| -mp		\| PGI \|\| pgcc, pgCC, pgfortran, etc. \|\| -mp
	\|-		\|-
	\| Intel \|\| icc, ifort \|\| -openmp		\| Intel \|\| icc, ifort \|\| -openmp
	\|-		\|-
	\| GNU \|\| gcc~~, gfortran~~, g++, ~~gcc44~~, etc. \|\| -fopenmp		\| GNU \|\| gcc, g++, gfortran, etc. \|\| -fopenmp
	\|-		\|-
	\|}		\|}

Shtsai: /* Sample job submission command on Sapelo2 */

2020-10-28T13:57:21Z

Sample job submission command on Sapelo2

← Older revision		Revision as of 09:57, 28 October 2020
Line 92:		Line 92:
	To submit a shared memory job that uses 4 OpenMP threads:		To submit a shared memory job that uses 4 OpenMP threads:
	<pre class="gcommand">		<pre class="gcommand">
	~~qsub~~ sub.sh		sbatch sub.sh
	</pre>		</pre>

	For more information about running jobs on Sapelo2, please see [[Running Jobs on Sapelo2]].		For more information about running jobs on Sapelo2, please see [[Running Jobs on Sapelo2]].

Shtsai at 13:57, 28 October 2020

2020-10-28T13:57:01Z

← Older revision		Revision as of 09:57, 28 October 2020
Line 1:		Line 1:
	[[Category:~~Sapelo2old~~]]		[[Category:Sapelo2]]

	===Compiling OpenMP code on Sapelo2===		===Compiling OpenMP code on Sapelo2===

Shtsai: /* Sample script to run a shared memory job using OpenMP */

2020-10-28T13:56:47Z

Sample script to run a shared memory job using OpenMP

← Older revision		Revision as of 09:56, 28 October 2020
Line 63:		Line 63:

	<pre class="gscript">		<pre class="gscript">
	#~~PBS -S~~ /bin/bash		#!/bin/bash
	#~~PBS~~ -~~q batch~~		#SBATCH --job-name=mctest # Job name
	#~~PBS~~ -~~N testjob~~		#SBATCH --partition=batch # Partition (queue) name
	#~~PBS~~ -~~l nodes~~=1~~:ppn~~=4		#SBATCH --ntasks=1 # Run a single task
	#~~PBS~~ -~~l walltime~~=48:00:00		#SBATCH --cpus-per-task=4 # Number of CPU cores per task
	#~~PBS~~ -~~l mem~~=~~30gb~~		#SBATCH --mem=4gb # Job memory request
	#~~PBS~~ -~~M username@uga~~.~~edu~~		#SBATCH --time=02:00:00 # Time limit hrs:min:sec
	#~~PBS -m abe~~		#SBATCH --output=mctest.%j.out # Standard output log
			#SBATCH --error=mctest.%j.err # Standard error log

	cd $~~PBS_O_WORKDIR~~		#SBATCH --mail-type=END,FAIL # Mail events (NONE, BEGIN, END, FAIL, ALL)
			#SBATCH --mail-user=username@uga.edu # Where to send mail

			cd $SLURM_SUBMIT_DIR

	export OMP_NUM_THREADS=4		export OMP_NUM_THREADS=4
	export OMP_PROC_BIND=true		export OMP_PROC_BIND=true

	~~echo~~		module load foss/2019b # load the appropriate module file, e.g. foss/2019b
	~~echo "Job ID: $PBS_JOBID"~~
	~~echo "Queue:~~ ~~$PBS_QUEUE"~~
	~~echo "Cores: $PBS_NP"~~
	~~echo "Nodes: $(cat $PBS_NODEFILE \| sort -u \| tr '\n' ' ')"~~
	~~echo~~

	time ./a.out > ~~outputfile~~		time ./a.out
			</pre>

	~~</pre>~~

	'''Please note''' that the number of cores requested by this job (~~ppn~~=4) is the same as the number of OpenMP threads that the application will use (export OMP_NUM_THREADS=4).		'''Please note''' that the number of cores requested by this job (--cpus-per-task=4) is the same as the number of OpenMP threads that the application will use (export OMP_NUM_THREADS=4).

	===Sample job submission command on Sapelo2===		===Sample job submission command on Sapelo2===

Shtsai at 16:05, 25 October 2020

2020-10-25T16:05:05Z

← Older revision		Revision as of 12:05, 25 October 2020
Line 1:		Line 1:
	[[Category:~~Sapelo2~~]]		[[Category:Sapelo2old]]

	===Compiling OpenMP code on Sapelo2===		===Compiling OpenMP code on Sapelo2===

Shtsai: /* Sample script to run a shared memory job using OpenMP */

2018-10-25T15:09:04Z

Sample script to run a shared memory job using OpenMP

← Older revision		Revision as of 11:09, 25 October 2018
Line 88:		Line 88:
	</pre>		</pre>

			'''Please note''' that the number of cores requested by this job (ppn=4) is the same as the number of OpenMP threads that the application will use (export OMP_NUM_THREADS=4).

	===Sample job submission command on Sapelo2===		===Sample job submission command on Sapelo2===