The DGX Software for Red Hat Enterprise Linux 7, EL7-19.08 update, is available. These updates are available only if you have enabled the optional update repository. See the document DGX-Software-Stack-for-Red-Hat-Enterprise-Linux-on-DGX <https://npncommunity.force.com/ESPCommunity/s/announcement/a4z1W000000WlMmQAK/dgx-software-stack-for-red-hat-enterprise-linux-on-dgx> (available to DGX customers with an NVIDIA Enterprise Support account) for instructions on updating the NVIDIA repositories.
The optional repository was created for updating the NVIDIA driver to the R418 driver branch, CUDA 10.1, and other software packages associated with CUDA 10.1.
Change Highlights
The following changes were made to the optional update repository.
Updated NVIDIA Driver to version 418.87: Includes security updates
Updated CUDA Toolkit to 10.1 Update 2 (10.1.243)
Updated cuDNN Library Runtime to 7.6.2
Updated DCGM to 1.6.6
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 <https://docs.nvidia.com/dgx/dgx-rhel-install-guide/index.html> for instructions.
Component |
Version |
|---|---|
GPU Driver |
410.104 |
NVIDIA System Health Monitor (NVSM) |
nvsm-cli 19.06.5-1 nvsm-dshm 19.06-2 nvsm-apis 19.06.9-1 nvhealth 19.06.8-1 |
Data Center GPU Management (DCGM) |
1.5.9 |
NCCL Runtime |
2.4.7+cuda10.0 |
cuDNN Library Runtime |
7.6.0.64-1+cuda10.0 |
TensorRT |
5.1.5+cuda10.0 |
CUDA Toolkit |
10.0-130 |
Component |
Version |
|---|---|
GPU Driver |
418.87 |
NVIDIA System Health Monitor (NVSM) |
Not included. Components are updated from the default repository. |
Data Center GPU Management (DCGM) |
1.6.6 |
NCCL Runtime |
2.4.7+cuda10.1 |
cuDNN Library Runtime |
7.6.2+cuda10.1 |
TensorRT |
5.1.5+cuda10.1 |
CUDA Toolkit |
10.1.243 |
Compatibility
NVIDIA has validated and tested the DGX Software version EL7-19.08 on the
NVIDIA DGX-2 with Red Hat Enterprise Linux 7.6 and CentOS, and
NVIDIA DGX-1 (Tesla V100) with Red Hat Enterprise Linux 7.6 and CentOS.
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
For installing on a fresh DGX system, see the DGX Software for Red Hat Enterprise Linux 7 - Installation Guide <https://docs.nvidia.com/dgx/dgx-rhel-install-guide/index.html> or the DGX Software for CentOS - Installation Guide <https://docs.nvidia.com/dgx/dgx-centos-install-guide/index.html>.
To obtain additional updates, issue the following.
sudo yum update
The updates will depend on which repositories you have enabled.
See the document `DGX-Software-Stack-for-Red-Hat-Enterprise-Linux-on-DGX <https://npncommunity.force.com/ESPCommunity/s/announcement/a4z1W000000WlMmQAK/dgx-software-stack-for-red-hat-enterprise-linux-on-dgx>` (available to DGX customers with an NVIDIA Enterprise Support account) for instructions on updating the NVIDIA repositories.
Known Issues
DGX-2: Ubuntu Appears as a Boot Option <unique_157798676>
DGX-1: DGX-1: DKMS May not Build for New Kernel During Driver Update<issue-dkms-build-during-driver-update>
DGX-1: NVSM Storage Alerts are Cleared When All Data Drives are Removed <unique_1914269314>
Black screen on BMC Remote Console with Red Hat Enterprise Linux 7.5
DGX-1: NVSM CLI Returns HTTP Code 500 Error After Hot-Plugging a Previously Removed SSD<issue-500-error-ssd-hot-plug>
DGX-1: Failure Reading Sector 0x0 May Occur on Reboot<issue-error-reading-sector-0x0-on-reboot>