The DGX Software for Red Hat Enterprise Linux 7, EL7-21.04 update, is available.
EL7-21.04 supports all DGX products - DGX A100, DGX-2, DGX-1 (with Tesla V100), DGX Station, and DGX Station A100.
Installing or updating to EL7-21.04 also updates the installed Red Hat Enterprise Linux 7 distribution to the latest version. If you require use of the Mellanox OpenFabrics Enterprise Distribution for Linux (MLNX_OFED), then before installing or updating to EL7-21.04, be sure that there is a MLNX_OFED package version available that supports the latest Red Hat Enterprise Linux 7 version.
To check the latest Red Hat Enterprise Linux 7 version, visit https://access.redhat.com/articles/3078
To check the MLNX_OFED package OS support, visit https://docs.mellanox.com/category/mlnxofedib, click the latest MLNX_OFED software version and then use the side menu to navigate to Release Notes->General Support in MLNX_OFED and view Supported Operating Systems.
If a supporting MLNX_OFED package has been released, then be sure to install it.
R450 Repository
The R450 repository must be enabled if you are installing on the DGX A100 or DGX Station A100. See the DGX Software for Red Hat Enterprise Linux Installation Guide for instructions.
Change Highlights
Added support for DGX Station A100.
The following changes were made in the Release 450 driver package.
Updated the GPU driver to 450.119.04
Updated NVSM to 20.09.21
Updated DCGM to 2.0.15
The following changes were made in the Release 418 driver package.
Updated the GPU driver to 418.197.02
Software Contents:
The following table provides version information for software included in the DGX Software Stack for Red Hat Enterprise Linux 7.
Unlike the DGX OS shipped with the NVIDIA DGX system, the DGX software stack for Red Hat does not include the Mellanox OpenFabrics Enterprise Distribution (MLNX_OFED) for Linux. This is due to the likelihood of the MLNX_OFED kernel being out of sync with the Red Hat distribution kernel. This can result in system instability. To use InfiniBand on the DGX system, see the DGX Software for Red Hat Enterprise Linux 7 Installation Guide for instructions.
Component |
Versions in the Release 418 Driver Package |
Versions in the Release 450 Driver Package |
|---|---|---|
GPU Driver |
418.197.02 |
450.119.04 |
NVIDIA System Management (NVSM) |
20.03.6 |
20.09.21 |
Data Center GPU Management (DCGM) |
1.7.4 |
2.0.15 |
DGX Station Theme |
dgxstation-desktop - 19.10-0 dgx-gnome - 19.10-0 |
dgxstation-desktop - 19.10-0 dgx-gnome - 19.10-0 (From R418 Repository) |
NCCL Runtime |
2.8.3+cuda10.1 |
2.8.3+cuda11.0 |
cuDNN Library Runtime |
8.0.5+cuda10.1 |
8.0.5+cuda11.0 |
cuBLAS |
10.2.2.214 |
11.1.0.229-1 |
CUDA Toolkit |
CUDA 10.1.243 |
CUDA 11.0.3 |
mpt3sas-dkms |
31.101.01.00 |
31.101.01.00 |
Compatibility
NVIDIA has validated and tested the DGX Software version EL7-21.04 with the following:
Linux Distribution and kernel
Red Hat Enterprise Linux 7.9
CentOS 7.9
Kernel 3.10.0-1160.25.1.el7
NVIDIA DGX systems
NVIDIA DGX A100
NVIDIA DGX-2
NVIDIA DGX-1 (V100)
NVIDIA DGX Station A100
NVIDIA DGX Station
MLNX OFED version 4.9-2.2.6.0
ConnectX Firmware
ConnectX-4: 12.28.2006
ConnectX-5: 16.28.2006
ConnectX-6: 20.28.2006
NVIDIA acknowledges the wide use of CentOS and understands that it is a community-developed derivative of the NVIDIA supported Red Hat Enterprise Linux. Support for CentOS is available directly from the CentOS community. NVIDIA ensures that NVIDIA provided software runs on tested CentOS versions and will try to identify and correct issues related to NVIDIA provided software.
Update Instructions
See the section Installing and Updating the Software install-update> for instructions.
Resolved Issues
Known Issues
DGX Station: The Symbolic Link to /usr/local/cuda Is Missing
DGX Station A100: nvidia-switch Error Message in DGX Station A100
Known Limitations
See Known Limitations for the list of known limitations and other issues that will not be fixed.