Home > Storage > PowerScale (Isilon) > Industry Solutions and Verticals > Analytics > PowerScale Deep Learning Infrastructure with NVIDIA DGX A100 Systems for Autonomous Driving > Hardware and software configuration
In this test, the hardware comprises four DGX A100 systems with four Isilon F800 nodes, and network switches. The following tables show the detailed hardware and software configurations.
Component | Number | Configuration |
Storage | 1 | Dell EMC Isilon F800 4-nodes cluster
|
Compute Node | 4 | NVIDIA DGX A100 system:
|
Storage Fabric Switch | 2 | NVIDIA Mellanox SN3700V Ethernet switches |
Compute Fabric Switch | 2 | NVIDIA Mellanox QM8700 InfiniBand HDR Switch |
Component | Version |
Dell EMC Isilon – OneFS |
|
NVIDIA Mellanox SN3700V – NCLU Version | 1.0-cl4.2.1u1 |
NVIDIA Mellanox SN3700V – Distribution Release | 4.2.1 |
NVIDIA Mellanox QM8700 Product Release | 3.9.0606 |
DGX A100 system – Base OS | 4.99.11 |
DGX A100 system – Linux kernel | 5.3.0-59-generic |
DGX A100 system – NVIDIA Data Center GPU Driver | 450.51.06 |
DGX A100 system – Ubuntu | 18.04.5 LTS |
NVIDIA NGC PyTorch Image | nvcr.io/nvidia/mxnet:20.06-py3 |