Controls/gitea-pages
0
2
Fork 0
Code Pull Requests Actions Activity

document Nvidia driver installation by ElRepo package or by manual installation

Browse Source
This commit is contained in:
buchel_k
2023-01-12 17:19:51 +01:00
parent 6293e088b6
commit cd869680a3
1 changed files with 11 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
rhel8/nvidia.md
View File

@@ -44,6 +44,13 @@ This is not recommended, still it is possible to do so by setting the exact driv
If the driver version is too old, it will install an older kernel version and you will need a second reboot to activate it.
### My hardware is too old
The oldest driver branch packaged by Nvidia for RHEL 8 is `470`. You might just live with the fallback to Nouveau (`nvidia::driver::enable: false` in Hiera) or you might try the drivers packaged by ElRepo: `nvidia::driver::branch: elrepo` (or when you want an specific ElRepo branch: `nvidia::driver::branch: 390xx`).
You might also just download and install the Nvidia driver manually.
Go to their [Download page](https://www.nvidia.de/Download/index.aspx), select and download the according installer and run it.
You best keep Puppet off your driver by setting `nvidia::driver::enable: false` in Hiera.
## Versioning Mess
@@ -64,11 +71,11 @@ Also it is not possible to find out from the package meta information if a drive
### What Driver \[Branch] for which Hardware
To figure out what driver branch to use for given hardware, go to their [Download page](https://www.nvidia.de/Download/index.aspx) and search its Linux driver. It will then point out a driver version and its first number points out the driver branch to use.
The most authorative way to do so is to chech the [Appendix A of the README of a decent driver](http://us.download.nvidia.com/XFree86/Linux-x86_64/525.78.01/README/supportedchips.html).
There search for your model or PCI ID. Then check out at the top of the respective table which legacy driver it still supports.
Or it might be the current driver.
Note that this is not always the full story. For example the Tesla K40c gives [driver 460.106.00](https://www.nvidia.de/Download/driverResults.aspx/182244/en-us), whereas the [470 driver](https://www.nvidia.com/Download/driverResults.aspx/194637/en-us/) still works, even though the hardware is not listed as supported there. My guess is that they somehow publickly differentiate between "Data Center Driver" and "Display Driver", but still they have everything in, or at least in the production/long term support branch.
Another option to figure out the driver is the third-party tool [`nvidia-detect`](http://elrepo.org/tiki/nvidia-detect) by ElRepo. It tells which driver package from ElRepo it suggests, but it can also be used to figure out which production/long term support branch can be used (and only production/long term support branches, e.g. it would never point out the 460 branch and this is how I figured out that Tesla K40c works with 470 despite the Nvidia documentation not saying so).
Another more automated option to figure out the driver is the third-party tool [`nvidia-detect`](http://elrepo.org/tiki/nvidia-detect) by ElRepo. It tells which driver package from ElRepo it suggests, but it can also be used to figure out which production/long term support branch can be used.
### CUDA - Driver Compatibility