GPU: Difference between revisions

From Research Computing Center Wiki
Jump to navigation Jump to search
Line 18: Line 18:
! scope="col" | Device memory
! scope="col" | Device memory
! scope="col" | GPU compute capability
! scope="col" | GPU compute capability
! scope="col" | CUDA version
! scope="col" | Minimum CUDA version
! scope="col" | Partition Name
! scope="col" | Partition Name
! scope="col" | Notes
! scope="col" | Notes
|-
|-
|-
|-
| 10 || 128 || 1TB  || Intel Sapphire Rapids || H100 || 4 || 80GB || 9.0 || >=11.8 || gpu_p, gpu_30d_p || Need to request --gres=gpu:H100, e.g.,
| 10 || 128 || 1TB  || Intel Sapphire Rapids || H100 || 4 || 80GB || 9.0 || 11.8 || gpu_p, gpu_30d_p || Need to request --gres=gpu:H100, e.g.,
<nowiki>#</nowiki>SBATCH --partition=gpu_p
<nowiki>#</nowiki>SBATCH --partition=gpu_p


Line 30: Line 30:
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
|-
|-
| 14 || 64 || 1TB  || AMD Milan || A100 || 4 || 80GB || 8.0 || >=11.0 || gpu_p, gpu_30d_p || Need to request --gres=gpu:A100, e.g.,
| 14 || 64 || 1TB  || AMD Milan || A100 || 4 || 80GB || 8.0 || 11.0 || gpu_p, gpu_30d_p || Need to request --gres=gpu:A100, e.g.,
<nowiki>#</nowiki>SBATCH --partition=gpu_p
<nowiki>#</nowiki>SBATCH --partition=gpu_p


Line 37: Line 37:
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
|-
|-
| 11 || 128 || 745GB  || AMD Genoa || L4 || 4 || 24GB || 8.9 || >=11.8 || gpu_p, gpu_30d_p || Need to request --gres=gpu:L4, e.g.,
| 11 || 128 || 745GB  || AMD Genoa || L4 || 4 || 24GB || 8.9 || 11.8 || gpu_p, gpu_30d_p || Need to request --gres=gpu:L4, e.g.,
<nowiki>#</nowiki>SBATCH --partition=gpu_p
<nowiki>#</nowiki>SBATCH --partition=gpu_p


Line 44: Line 44:
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
|-
|-
| 2 || 32 || 192GB || Intel Skylake || P100 || 1 || 16GB || 6.0 || >=8.0|| gpu_p, gpu_30d_p ||Need to request --gres=gpu:P100, e.g.,
| 2 || 32 || 192GB || Intel Skylake || P100 || 1 || 16GB || 6.0 || 8.0|| gpu_p, gpu_30d_p ||Need to request --gres=gpu:P100, e.g.,
<nowiki>#</nowiki>SBATCH --partition=gpu_p
<nowiki>#</nowiki>SBATCH --partition=gpu_p


Line 51: Line 51:
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
<nowiki>#</nowiki>SBATCH --time=7-00:00:00  
|-
|-
| 1 || 64 || 1TB  || AMD Milan || A100 || 4 || 80GB || 8.0 || >=11.0 || buyin partition || rowspan="8" | Available on '''batch''' for all users up to '''4 hours''', e.g.,
| 1 || 64 || 1TB  || AMD Milan || A100 || 4 || 80GB || 8.0 || 11.0 || buyin partition || rowspan="8" | Available on '''batch''' for all users up to '''4 hours''', e.g.,
<nowiki>#</nowiki>SBATCH --partition=batch
<nowiki>#</nowiki>SBATCH --partition=batch


<nowiki>#</nowiki>SBATCH --gres=gpu:A100:1 or  
<nowiki>#</nowiki>SBATCH --gres=gpu:A100:1 or  
<nowiki>#</nowiki>SBATCH --gres=gpu:L4:1 or


<nowiki>#</nowiki>SBATCH --gres=gpu:V100:1 or  
<nowiki>#</nowiki>SBATCH --gres=gpu:V100:1 or  


<nowiki>#</nowiki>SBATCH --gres=gpu:V100S:1 or
<nowiki>#</nowiki>SBATCH --gres=gpu:V100S:1
 
<nowiki>#</nowiki>SBATCH --gres=gpu:L4:1


<nowiki>#</nowiki>SBATCH --time=4:00:00
<nowiki>#</nowiki>SBATCH --time=4:00:00
|-
|-
| 2 || 28 || 192GB || Intel Skylake || V100 || 1 || 16GB || 7.0|| >=9.0 || buyin partition  
| 2 || 64 || 745GB  || AMD Genoa || L4 || 4 || 24GB || 8.9 || 11.8 || buyin partition
|-
|-
| 2 || 32 || 192GB || Intel Skylake || V100 || 1 || 16GB || 7.0 || >=9.0 || buyin partition  
| 2 || 28 || 192GB || Intel Skylake || V100 || 1 || 16GB || 7.0|| 9.0 || buyin partition  
|-
|-
| 2 || 32 || 384GB || Intel Skylake || V100 || 1 || 32GB || 7.0 || >=9.0 || buyin partition  
| 2 || 32 || 192GB || Intel Skylake || V100 || 1 || 16GB || 7.0 || 9.0 || buyin partition  
|-
|-
| 2 || 64 || 128GB || AMD Naples || V100 || 2 || 32GB || 7.0 || >=9.0 || buyin partition  
| 2 || 32 || 384GB || Intel Skylake || V100 || 1 || 32GB || 7.0 || 9.0 || buyin partition  
|-
|-
| 1 || 64 || 128GB || AMD Naples || V100 || 1 || 32GB || 7.0 || >=9.0 || buyin partition  
| 2 || 64 || 128GB || AMD Naples || V100 || 2 || 32GB || 7.0 || 9.0 || buyin partition  
|-
|-
| 4 || 64 || 128GB || AMD Rome || V100S || 1 || 32GB || 7.0 || >=9.0 || buyin partition  
| 1 || 64 || 128GB || AMD Naples || V100 || 1 || 32GB || 7.0 || 9.0 || buyin partition  
|-
|-
| 2 || 64 || 745GB  || AMD Genoa || L4 || 4 || 24GB || 8.9 || >=11.8 || buyin partition
| 4 || 64 || 128GB || AMD Rome || V100S || 1 || 32GB || 7.0 || 9.0 || buyin partition  
|-
|-
|}
|}

Revision as of 16:33, 31 August 2024


GPU Computing on Sapelo2

Hardware

For a description of the Graphics Processing Units (GPU) device specifications, please see GPU Hardware.

The following table summarizes the GPU devices available on sapelo2:

Number of nodes CPU cores per node Host memory per node CPU processor GPU model GPU devices per node Device memory GPU compute capability Minimum CUDA version Partition Name Notes
10 128 1TB Intel Sapphire Rapids H100 4 80GB 9.0 11.8 gpu_p, gpu_30d_p Need to request --gres=gpu:H100, e.g.,

#SBATCH --partition=gpu_p

#SBATCH --gres=gpu:H100:1

#SBATCH --time=7-00:00:00

14 64 1TB AMD Milan A100 4 80GB 8.0 11.0 gpu_p, gpu_30d_p Need to request --gres=gpu:A100, e.g.,

#SBATCH --partition=gpu_p

#SBATCH --gres=gpu:A100:1

#SBATCH --time=7-00:00:00

11 128 745GB AMD Genoa L4 4 24GB 8.9 11.8 gpu_p, gpu_30d_p Need to request --gres=gpu:L4, e.g.,

#SBATCH --partition=gpu_p

#SBATCH --gres=gpu:L4:1

#SBATCH --time=7-00:00:00

2 32 192GB Intel Skylake P100 1 16GB 6.0 8.0 gpu_p, gpu_30d_p Need to request --gres=gpu:P100, e.g.,

#SBATCH --partition=gpu_p

#SBATCH --gres=gpu:P100:1

#SBATCH --time=7-00:00:00

1 64 1TB AMD Milan A100 4 80GB 8.0 11.0 buyin partition Available on batch for all users up to 4 hours, e.g.,

#SBATCH --partition=batch

#SBATCH --gres=gpu:A100:1 or

#SBATCH --gres=gpu:L4:1 or

#SBATCH --gres=gpu:V100:1 or

#SBATCH --gres=gpu:V100S:1

#SBATCH --time=4:00:00

2 64 745GB AMD Genoa L4 4 24GB 8.9 11.8 buyin partition
2 28 192GB Intel Skylake V100 1 16GB 7.0 9.0 buyin partition
2 32 192GB Intel Skylake V100 1 16GB 7.0 9.0 buyin partition
2 32 384GB Intel Skylake V100 1 32GB 7.0 9.0 buyin partition
2 64 128GB AMD Naples V100 2 32GB 7.0 9.0 buyin partition
1 64 128GB AMD Naples V100 1 32GB 7.0 9.0 buyin partition
4 64 128GB AMD Rome V100S 1 32GB 7.0 9.0 buyin partition

Software

Sapelo2 has the following tools for programming for GPUs:

1. NVIDIA CUDA toolkit

Several versions of the CUDA toolkit are available. Please see our CUDA page.

2. PGI/CUDA compilers

The PGI compilers available on Sapelo2 support GPU acceleration, including Fortran/CUDA.

For more information on the GPU support of PGI compilers, please visit the PGI website http://www.pgroup.com/resources/cudafortran.htm

For information on versions of PGI compilers installed on Sapelo2, please see Code Compilation on Sapelo2.

3. OpenACC

Using the NVIDIA HPC SDK compiler suite or the old PGI Accelerator compilers, programmers can accelerate applications on x64+accelerator platforms by adding OpenACC compiler directives to Fortran and C programs and then recompiling with appropriate compiler options. Please see https://developer.nvidia.com/hpc-sdk and http://www.pgroup.com/resources/accel.htm

OpenACC is also supported by GNU compilers, especially the latest versions, e.g. GNU 7.2.0, installed on Sapelo2. For more information on OpenACC support by GNU compilers, please refer to https://gcc.gnu.org/wiki/OpenACC

For information on versions of GNU compilers installed on Sapelo2, please see Code Compilation on Sapelo2.

Running Jobs

For information on how to run GPU jobs on Sapelo2, please refer to Running Jobs on Sapelo2.

Retrieved from "https://wiki.gacrc.uga.edu/index.php?title=GPU&oldid=22029"