Cuda
module avail cuda/
CUDA (Compute Unified Device Architecture) libraries are software libraries developed by NVIDIA for parallel computing on NVIDIA GPUs (Graphics Processing Units).
These libraries provide optimized functions and algorithms that leverage the parallel processing power of GPUs for various computational tasks.
License
To use Cuda, you will also need to accept the license for cuDNN library.
Usage
GPU clusters
| Cluster | Nodes | GPUs per node | Compute capability |
Mem [MiB] |
CuDNN |
|---|---|---|---|---|---|
| adan.grid.cesnet.cz | adan[1-61].grid.cesnet.cz | 2x Tesla T4 | 7.5 | 15 109 | YES |
| bee.cerit-sc.cz | bee[1-10].cerit-sc.cz | 2x H100 NVL | 9.0 | 95 830 | YES |
| cha.natur.cuni.cz | cha.natur.cuni.cz | 8x GeForce RTX 2080 Ti | 7.5 | 11 019 | YES |
| fau.natur.cuni.cz | fau[1-3].natur.cuni.cz | 8x Quadro RTX 5000 | 7.5 | 16 125 | YES |
| fer.natur.cuni.cz | fer[1-3].natur.cuni.cz | 8x RTX A4000 | 8.6 | 16 117 | YES |
| galdor.metacentrum.cz | galdor[1-20].metacentrum.cz | 4x A40 | 8.6 | 45 634 | YES |
| gita.cerit-sc.cz | gita[1-7].cerit-sc.cz | 2x GeForce RTX 2080 Ti | 7.5 | 11 019 | YES |
| glados.cerit-sc.cz | glados1.cerit-sc.cz | 1x TITAN V GPU | 7.0 | 12 066 | YES |
| glados.cerit-sc.cz | glados[2-7].cerit-sc.cz | 2x GeForce RTX 2080 | 7.5 | 7 982 | YES |
| glados.cerit-sc.cz | glados[10-13].cerit-sc.cz | 2x 1080Ti GPU | 6.1 | 11 178 | YES |
| grimbold.metacentrum.cz | grimbold.metacentrum.cz | 2x Tesla P100 | 6.0 | 12 198 | YES |
| konos.fav.zcu.cz | konos[1-8].fav.zcu.cz | 4x GeForce GTX 1080 Ti | 6.1 | 11 178 | YES |
| luna2022.fzu.cz | luna[201-206].fzu.cz | 1x A40 | 8.6 | 45 634 | YES |
| zia.cerit-sc.cz | zia[1-5].cerit-sc.cz | 4x A100 | 8.0 | 40 536 | YES |
GPU jobs
- GPU queues: gpu (24 hours max) and gpu_long (up to 336 hours), both with open access for all MetaCentrum members
- GPU jobs on the konos cluster can be also run via the priority queue iti (queue for users from ITI - Institute of Theoretical Informatics, Univ. of West Bohemia)
- zubat cluster is available for any job which will run 24 hours at most.
- Users from CEITEC MU and NCBR can run jobs via privileged queues on the zubat cluster.
- The current version of the cuda drivers (parameter
cuda_version) can be verified interactively in the qsub command assembler.
Requesting GPUs
The key scheduling constraint is to prevent jobs from sharing GPUs. To ensure this always use the gpu=X flag in qsub and request one of the gpu queues (gpu, gpu_long, iti).
qsub -l select=1:ncpus=1:mem=10gb:ngpus=X -q gpu
where X means a number of GPU cards required. By default
resources_default.gpu=1
If a job requires more GPU cards than it asks (or is available), prolog does not run it.
To plan your job on clusters with certain compute capability, use qsub command like this:
qsub -q gpu -l select=1:ncpus=1:ngpus=X:gpu_cap=cuda35 <job batch file>
Using the PBS parameter gpu_mem is possible to specify the minimum amount of memory that the GPU card will have.
qsub -q gpu -l select=1:ncpus=1:ngpus=1:gpu_mem=10gb ...
Example
qsub -I -q gpu -l select=1:ncpus=1:ngpus=1:scratch_local=10gb:gpu_mem=10gb -l walltime=24:0:0
Interactive job requests 1 machine, 1 CPU and 1 GPU card for 24 hours.
FAQs
Q: How can I recognize which GPUs are reserved for me by planning system?
A: IDs of GPU cards are stored in CUDA_VISIBLE_DEVICES variable. These IDs are mapped to CUDA tools virtual IDs. Though if CUDA_VISIBLE_DEVICES contains value 2, 3 then CUDA tools will report IDs 0, 1.
Q: I want to use the NVIDIA CuDNN library, which GPU clusters do support it?
A: Those which have GPU with compute capability > 3.0, which means all clusters (see the table above)