1. OIT HPC
  2. Running Jobs
  3. Deployment Status

Slurm Deployment Status

Current status of Slurm testing and deployment.

LSF to Slurm Migration 2026 Timeline: Feb - HPC Committee Approval; Apr-May - Early and General User Testing; Jun-Aug - Partner Hardware Moves and Production Ramp; Aug-Dec - Production Operation; Dec 16-17 - All Remaining Nodes Moved, LSF Retired
Testing Phase: Slurm is currently in testing with limited hardware. Production deployment will follow successful testing.

Current Test Environment

The following resources are currently available for Slurm testing. Nodes are actively being transitioned from LSF to Slurm. In addition all new partner nodes are being deployed to Slurm partitions. Therefore, this is a constantly evolving list.

Compute nodes are either available via LSF or via Slurm. Compute nodes cannot be shared between both schedulers. Partner nodes (and partner LSF queue) can be moved by partner request. Note some partner owned nodes are already being moved with goal of having about half the compute nodes managed by Slurm and about half still managed by LSF for Fall Semester. Additional nodes will be shifted during the semester is Slurm usage warrents that. No nodes will be moved back to LSF. If LSF queue wait times are long please shift your work to Slurm.

Compute Nodes

Node Type CPU Architecture Quantity Constraint
CPU Node Intel Xeon E5 v3 (Haswell) 45 --constraint=haswell
CPU Node Intel Xeon E5 v4 (Broadwell) 35 --constraint=broadwell
CPU Node Intel 1st Generation Xeon SP (Skylake) 81 --constraint=skylake
CPU Node Intel 2nd Generation Xeon SP (Cascade Lake) 21 --constraint=cascadelake
CPU Node Intel 3rd Generation Xeon SP (Sapphire Rapids) 2 --constraint=sapphirerapids
CPU Node AMD 4th Generation EPYC (Genoa) 7 --constraint=genoa
CPU Node AMD 5th Generation EPYC (Turin) 1 --constraint=turin

GPU Nodes

Node Type GPUs Quantity Request
GPU Node 2x NVIDIA GTX 1080 (8 GB each) 1 --partition=gpu --gres=gpu:gtx_1080:N
GPU Node 4x NVIDIA RTX 2080 (8 GB each) 1 --partition=gpu --gres=gpu:rtx_2080:N
GPU Node 2x NVIDIA P100 (16 GB each) 1 --partition=gpu --gres=gpu:p100:N
GPU Node 2x NVIDIA A10 (24 GB each) 2 --partition=gpu --gres=gpu:a10:N
GPU Node 2x NVIDIA A30 (24 GB each) 4 --partition=gpu --gres=gpu:a30:N
GPU Node 4x NVIDIA A100 (40 GB each) 1 --partition=gpu --gres=gpu:a100:N
GPU Node 4x NVIDIA L40 (48 GB each) 1 --partition=gpu --gres=gpu:l40:N
GPU Node 4x NVIDIA L40S (48 GB each) 1 --partition=gpu --gres=gpu:l40s:N

MPI Modules with Slurm Integration

The following modules provide MPI compilers with Slurm integration, allowing use of srun in place of mpirun:

Environment Compiler MPI Library Module
GNU + OpenMPI GCC 11.5 OpenMPI 4.1.8 openmpi-gcc/openmpi4.1.8-gcc11.5.0-slurm
Intel + Intel MPI Intel 2025.3 Intel MPI PrgEnv-intel/2025.3-slurm
NVIDIA HPC SDK NVHPC 26.1 OpenMPI PrgEnv-nvidia/26.1-slurm

These modules are configured for Slurm's Process Management Interface (PMI), enabling efficient job launch with srun. See Parallel Jobs Guide for usage details.

Test Job Examples

CPU job on Haswell node

#!/bin/bash
#SBATCH --job-name=test_haswell
#SBATCH --output=test.out.%j
#SBATCH --ntasks=4
#SBATCH --constraint=haswell
#SBATCH --mem=4G
#SBATCH --time=00:30:00

hostname
echo "Running on Haswell"

CPU job on Broadwell node

#!/bin/bash
#SBATCH --job-name=test_broadwell
#SBATCH --output=test.out.%j
#SBATCH --ntasks=4
#SBATCH --constraint=broadwell
#SBATCH --mem=8G
#SBATCH --time=00:30:00

hostname
echo "Running on Broadwell"

CPU job on Genoa node

#!/bin/bash
#SBATCH --job-name=test_genoa
#SBATCH --output=test.out.%j
#SBATCH --ntasks=48
#SBATCH --partition=compute_partners
#SBATCH --constraint=genoa
#SBATCH --mem=16G
#SBATCH --time=00:30:00

hostname
echo "Running on Genoa"

GPU job

#!/bin/bash
#SBATCH --job-name=test_gpu
#SBATCH --output=test_gpu.out.%j
#SBATCH --partition=gpu
#SBATCH --gres=gpu:rtx_2080:1
#SBATCH --ntasks=1
#SBATCH --mem=16G
#SBATCH --time=00:30:00

module load cuda
nvidia-smi

Checking Availability

View the current state of test nodes: 

# Show all nodes managed by Slurm
si --nodes --slurm

# Show GPU availability
si --nodes --gpu

Known Limitations

  • Testing environment - configurations may change without notice

Reporting Issues

If you encounter issues during testing, please report them to HPC support with:

  • Job ID
  • Batch script used
  • Error messages or unexpected behavior
  • Output from squeue -j JOBID or sacct -j JOBID

Documentation

While testing with limited resources, you can familiarize yourself with Slurm using our documentation: