Using GPUs with Slurm

For general advice on job scheduling, see Running jobs.

Available hardware

These are the GPUs currently available:

Cluster	# of Nodes	Slurm type specifier	Per node			GPU model	Compute Capability	GPU mem (GiB)	Notes
			CPU cores	CPU memory	GPUs
Béluga	172	v100	40	191000M	4	V100-SXM2	70	16	All GPUs associated with the same CPU socket, connected via NVLink
Cedar	114	p100	24	128000M	4	P100-PCIE	60	12	Two GPUs per CPU socket
	32	p100l	24	257000M	4	P100-PCIE	60	16	All GPUs associated with the same CPU socket
	192	v100l	32	192000M	4	V100-SXM2	70	32	Two GPUs per CPU socket; all GPUs connected via NVLink
Graham	160	p100	32	127518M	2	P100-PCIE	60	12	One GPU per CPU socket
	7	v100	28	183105M	8	V100-PCIE	70	16	See Graham: Volta GPU nodes
	2	v100l	28	183105M	8	V100-?	70	32	See Graham: Volta GPU nodes
	30	t4	44	192000M	4	Tesla T4	75	16	Two GPUs per CPU socket
	6	t4	16	192000M	4	Tesla T4	75	16
Hélios	15	k20	20	110000M	8	K20	35	5	Four GPUs per CPU socket
	6	k80	24	257000M	16	K80	37	12	Eight GPUs per CPU socket
Mist	54	(none)	32	256GiB	4	V100-SXM2	70	32	See Mist specifications
Narval	159	a100	48	510000M	4	A100	80	40	Two GPUs per CPU socket; all GPUs connected via NVLink
Arbutus	Cloud resources are not schedulable via Slurm. See Cloud resources for details of available hardware.

Specifying the type of GPU to use

Some clusters have more than one GPU type available (Cedar, Graham, Hélios), and some clusters only have GPUs on certain nodes (Béluga, Cedar, Graham). You can choose the type of GPU to use by supplying to Slurm the type specifier given in the table above, e.g.:

#SBATCH --gres=gpu:p100:1

If you do not supply a type specifier, Slurm may send your job to a node equipped with any type of GPU. For certain workflows this may be undesirable. For example, molecular dynamics code requires high double-precision performance, and therefore T4 GPUs are not appropriate. In such a case, make sure you include a type specifier.

Mist

Mist is a cluster comprised of IBM Power9 CPUs (not Intel x86!) and NVIDIA V100 GPUs. Users with access to Niagara can also access Mist. To specify job requirements on Mist, please see the specific instructions on the SciNet web site.

Single-core job

If you need only a single CPU core and one GPU:

File : gpu_serial_job.sh

#!/bin/bash
#SBATCH --account=def-someuser
#SBATCH --gres=gpu:1              # Number of GPUs (per node)
#SBATCH --mem=4000M               # memory (per node)
#SBATCH --time=0-03:00            # time (DD-HH:MM)
./program                         # you can use 'nvidia-smi' for a test

Multi-threaded job

For GPU jobs asking for multiple CPUs in a single node:

File : gpu_threaded_job.sh

#!/bin/bash
#SBATCH --account=def-someuser
#SBATCH --gres=gpu:1              # Number of GPU(s) per node
#SBATCH --cpus-per-task=6         # CPU cores/threads
#SBATCH --mem=4000M               # memory per node
#SBATCH --time=0-03:00            # time (DD-HH:MM)
export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK
./program

For each GPU requested on:

• Béluga, we recommend no more than 10 CPU cores.
• Cedar, we recommend no more than 6 CPU cores per P100 GPU (p100 and p100l) and no more than 8 CPU cores per V100 GPU (v100l).
• Graham, we recommend no more than 16 CPU cores.

MPI job

File : gpu_mpi_job.sh

#!/bin/bash
#SBATCH --account=def-someuser
#SBATCH --gres=gpu:4              # Number of GPUs per node
#SBATCH --nodes=2                 # Number of nodes
#SBATCH --ntasks=48               # Number of MPI process
#SBATCH --cpus-per-task=1         # CPU cores per MPI process
#SBATCH --mem=120G                # memory per node
#SBATCH --time=0-03:00            # time (DD-HH:MM)
export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK
srun ./program

Whole nodes

If your application can efficiently use an entire node and its associated GPUs, you will probably experience shorter wait times if you ask Slurm for a whole node. Use one of the following job scripts as a template.

Requesting a GPU node on Graham

File : graham_gpu_node_job.sh

#!/bin/bash
#SBATCH --nodes=1
#SBATCH --gres=gpu:2
#SBATCH --ntasks-per-node=32
#SBATCH --mem=127000M
#SBATCH --time=3:00
#SBATCH --account=def-someuser
nvidia-smi

Requesting a P100 GPU node on Cedar

File : cedar_gpu_node_job.sh

#!/bin/bash
#SBATCH --nodes=1
#SBATCH --gres=gpu:p100:4
#SBATCH --ntasks-per-node=24
#SBATCH --exclusive
#SBATCH --mem=125G
#SBATCH --time=3:00
#SBATCH --account=def-someuser
nvidia-smi

Requesting a P100-16G GPU node on Cedar

There is a special group of GPU nodes on Cedar which have four Tesla P100 16GB cards each. (Other P100 GPUs in the cluster have 12GB and the V100 GPUs have 32G.) The GPUs in a P100L node all use the same PCI switch, so the inter-GPU communication latency is lower, but bandwidth between CPU and GPU is lower than on the regular GPU nodes. The nodes also have 256GB RAM. You may only request these nodes as whole nodes, therefore you must specify --gres=gpu:p100l:4. P100L GPU jobs up to 28 days can be run on Cedar.

File : p100l_gpu_job.sh

#!/bin/bash
#SBATCH --nodes=1 
#SBATCH --gres=gpu:p100l:4   
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=24    # There are 24 CPU cores on P100 Cedar GPU nodes
#SBATCH --mem=0               # Request the full memory of the node
#SBATCH --time=3:00
#SBATCH --account=def-someuser
hostname
nvidia-smi

Packing single-GPU jobs within one SLURM job

If you need to run four single-GPU programs or two 2-GPU programs for longer than 24 hours, GNU Parallel is recommended. A simple example is given below:

cat params.input | parallel -j4 'CUDA_VISIBLE_DEVICES=$(({%} - 1)) python {} &> {#}.out'

In this example the GPU ID is calculated by subtracting 1 from the slot ID {%}. {#} is the job ID, starting from 1.

A params.input file should include input parameters in each line, like this:

code1.py
code2.py
code3.py
code4.py
...

With this method, users can run multiple tasks in one submission. The -j4 parameter means GNU Parallel can run a maximum of four concurrent tasks, launching another as soon as each one ends. CUDA_VISIBLE_DEVICES is used to ensure that two tasks do not try to use the same GPU at the same time.

CUDA Compute Capability

When you are compiling CUDA code on clusters it’s important to know which is the Compute Capability of the GPU that you are targeting. If you get the following error during the compile time:

nvcc fatal : Unsupported gpu architecture 'compute_XX'

or this error during running your CUDA code on a compute node with GPU:

no kernel image is available for execution on the device (209)

you can fix it by adding the correct FLAG to “nvcc” call:

-gencode arch=compute_XX,code=[sm_XX,compute_XX]

or if you are using CMake to build your project, by providing the following flag:

cmake .. -DCMAKE_CUDA_ARCHITECTURES=XX

where “XX” is the Compute Capability of the Nvidia GPU board that you are going to use. Now you need to know the correct value to replace “XX“, you can find it under Compute Capability column on the above table.

For example, if you are running your code on a Narval A100 node, you find that its Compute Capability is 80, so the correct FLAG to use in the compiler is

-gencode arch=compute_80,code=[sm_80,compute_80]

or the following command to configure CMake:

cmake .. -DCMAKE_CUDA_ARCHITECTURES=80

Message Passing Interface