Skip to content

GPU Resources on HiPerGator

GPU Group Allocations

Normalized Graphics Processor Units (NGUs) represent general purpose computational GPU (GPGPU) units on HiPerGator. Each NGU is equivalent to 1 GPU at this time, but this may be changed in the future to optimize the GPU use.

In order to use GPU resources HiPerGator groups need to have an active NGU allocation.

Check if a group has GPUs allocated and available with the command slurmInfo -g group_name (the command is in the "ufrc" environment module).

GPU Hardware on HiPerGator

We have the following types of NVIDIA GPU nodes available:

Node Specifications NVIDIA L4 NVIDIA RTX PRO 6000 NVIDIA B200
Host Quantity 200 45 31
Host Architecture AMD TURIN AMD EPYC 9555 Intel Xeon 8570
Host Memory 753 GB 1,536 GB 2 TB
Host Interconnect ConnectX-7 IB ConnectX-7 IB ConnectX-7 IB
CPU cores per Host 96 128 112
CPU cores per Socket 96 64 56
GPUs per Host 3 8 8
CPU cores per GPU 4 (reserved per GPU) 16 14
Memory per GPU 24 GB 96 GB 180 GB
Slurm partition hpg-turin,hwgui,gpu hpg-rtx6000 hpg-b200
Slurm Feature l4 rtx-pro-6000 b200
GRES TYPE l4 rtx-pro-6000 b200
Technical Ref Specifications Specifications Specifications

Legend:

  • Slurm features are used in '--constraint' job specifications
  • GRES types are used in '--gres=gpu:TYPE:NUMBER' job specifications

For a list of additional node features, see the Available Node Features page.

GPU Partitions on HiPerGator

  • hpg-b200: Partition representing the NVIDIA DGX B200 SuperPod.
  • hpg-rtx6000: Lenovo SR675v3 based partition with NVIDIA RTX PRO 6000 GPUs.
  • hpg-turin: Regular HPG4 nodes integrated with 3 NVIDIA L4 cards per node. This partition does NOT need to be specified in a job script. It will be selected automatically if a GPU is requested.
  • hwgui: Hardware accelerated GPU partition for visualization applications with NVIDIA L4 cards.

Usage Documentation