From 34bfb0326ae1d937b8d18796c1c0c9211ff56fe1 Mon Sep 17 00:00:00 2001
From: Jannis Klinkenberg <j.klinkenberg@itc.rwth-aachen.de>
Date: Wed, 19 Mar 2025 15:07:06 +0100
Subject: [PATCH] filled generic job README.md
---
generic-job-scripts/README.md | 19 +++++++++++++++++--
.../openmp_multi-threading_job.sh | 5 +++--
2 files changed, 20 insertions(+), 4 deletions(-)
diff --git a/generic-job-scripts/README.md b/generic-job-scripts/README.md
index f7da021..1763d6a 100644
--- a/generic-job-scripts/README.md
+++ b/generic-job-scripts/README.md
@@ -1,3 +1,18 @@
-# Generic SLURM Job Scripts
+# Generic Slurm Job Script Examples
-... coming soon ...
+This folder contains common job script examples and best practices. You can submit jobs to the Slurm batch system via `sbatch <script-name>.sh`.
+
+## What can you find here?
+
+| File/Folder | Description |
+|--------|-------------|
+| [beeond_job.sh](beeond_job.sh) | Job script for setting up and using BeeOND (BeeGFS On Demand) in an HPC environment. |
+| [gpu_job_1gpu.sh](gpu_job_1gpu.sh) | Runs a job with 1 GPU and a single process. |
+| [gpu_job_2gpus-1proc.sh](gpu_job_2gpus-1proc.sh) | Runs a job with 2 GPUs and a single process. Useful for tasks that require multi-GPU acceleration but not multi-processing. |
+| [gpu_job_2gpus-2procs.sh](gpu_job_2gpus-2procs.sh) | Runs a job with 2 GPUs and and 2 separate processes. Commonly used for parallel deep learning training. |
+| [gpu_job_4gpus-4procs.sh](gpu_job_4gpus-4procs.sh) | Runs a job with 4 GPUs and and 4 separate processes (full node with 4x H100). Commonly used for parallel deep learning training. |
+| [gpu_job_8gpus-8procs.sh](gpu_job_8gpus-8procs.sh) | Runs a job with 8 GPUs and and 8 separate processes (2 full nodes with 4x H100). Commonly used for parallel deep learning training. |
+| [hybrid_mpi_openmp_job.sh](hybrid_mpi_openmp_job.sh) | Hybrid job script combining MPI (distributed computing) with OpenMP (shared-memory parallelism). Ideal for hybrid HPC workloads. |
+| [mpi_job_basic.sh](mpi_job_basic.sh) | A basic MPI job script, useful for testing and learning MPI-based job submission. |
+| [mpi_job_1node.sh](mpi_job_1node.sh) | Runs an MPI job on a single node, demonstrating intra-node parallel processing with multiple processes per node. |
+| [mpi_job_2nodes.sh](mpi_job_2nodes.sh) | Runs an MPI job spanning 2 full compute nodes, demonstrating inter-node parallelism and distributed computing across multiple machines. |
diff --git a/generic-job-scripts/openmp_multi-threading_job.sh b/generic-job-scripts/openmp_multi-threading_job.sh
index 0eda131..d5bed48 100644
--- a/generic-job-scripts/openmp_multi-threading_job.sh
+++ b/generic-job-scripts/openmp_multi-threading_job.sh
@@ -5,7 +5,7 @@
#SBATCH --nodes=1 # pure multi-threading restricted to single node
#SBATCH --ntasks-per-node=1 # pure multi-threading restricted to single process
-#SBATCH --cpus-per-task=48 # request desired number of CPU cores or threads per process (default: 1)
+#SBATCH --cpus-per-task=96 # request desired number of CPU cores or threads per process (default: 1). Here, we want to request all CPU cores
# Note: available main memory is also scaling with
# number of cores if not specified otherwise
#SBATCH --time=00:15:00 # max. run time of the job
@@ -26,8 +26,9 @@ echo "Current machine: $(hostname)"
# Example: OpenMP
-# set number of threads
+# set number of OpenMP threads to be used
export OMP_NUM_THREADS=${SLURM_CPUS_PER_TASK} # usually automatically set by SLURM
+# Note: you can also use less cores/threads or experiment with different number of cores/threads in the same job
# enable thread binding to pyhsical CPU cores
export OMP_PLACES=cores
--
GitLab