What driver is needed to use the NVIDIA GPU Computing Processors with Red Hat Enterprise Linux or Red Hat Enterprise Linux OpenStack Platform?

Solution Verified - Updated -

Environment

  • Red Hat Enterprise Linux 4
  • Red Hat Enterprise Linux 5
  • Red Hat Enterprise Linux 6
  • Red Hat Enterprise Linux 7
  • Red Hat Enterprise Linux 8
  • Red Hat Enterprise Linux OpenStack Platform 3
  • Red Hat Enterprise Linux OpenStack Platform 4
  • Red Hat Enterprise Linux OpenStack Platform 5
  • Red Hat Enterprise Linux OpenStack Platform 6
  • Red Hat Enterprise Linux OpenStack Platform 7
  • Red Hat OpenStack Platform 8
  • Red Hat OpenStack Platform 9
  • Red Hat OpenStack Platform 10
  • Red Hat OpenStack Platform 11
  • Red Hat OpenStack Platform 12
  • Red Hat OpenStack Platform 13
  • Red Hat OpenStack Platform 14
  • Red Hat OpenStack Platform 15

  • NVIDIA Tesla C870 Computing Processor

  • NVIDIA Tesla C10xx Computing Processor
  • NVIDIA Tesla C20xx Computing Processor
  • NVIDIA Tesla M10xx Computing Processor
  • NVIDIA Tesla M20xx Computing Processor
  • NVIDIA Tesla Kxx Computing Processor
  • NVIDIA Tesla Pxx Computing Processor
  • NVIDIA Tesla Vxx Computing Processor
  • NVIDIA GRID Kx Computing card
  • All other unlisted NVIDIA GPU Computing Processors

Issue

A dedicated driver is required to use the full processing capabilities of the NVIDIA GPU cards - including but not limited to the Tesla C870 and C10xx, C20xx, M10xx, M20xx, Kxx, Pxx Computing Processor cards with Red Hat Enterprise Linux or Red Hat Enterprise Linux OpenStack Platform. The Tesla and other NVIDIA compute processing cards are generally referred to as dedicated general purpose GPU's or GPGPU http://en.wikipedia.org/wiki/GPGPU. Red Hat is currently unable to ship this driver as it is not open source.

Resolution

The driver for these processors is included within the NVIDIA binary driver package, which can be downloaded directly from the NVIDIA website at http://www.nvidia.com

While Red Hat is unable to supply these drivers directly we will fully support the installation on which these drivers are used, and will work with NVIDIA to resolve any problems found within the Red Hat code. Support for third party applications developed around the NVIDIA GPU and dedicated driver is not implied.

There is preliminary support for some of these GPGPU's with the open source nouveau http://nouveau.freedesktop.org/wiki/ driver included with RHEL and RHELOSP. You can see a feature matrix here http://nouveau.freedesktop.org/wiki/FeatureMatrix/. How the code names map to official names can be seen here http://nouveau.freedesktop.org/wiki/CodeNames/. Keep in mind that the implemented features outlined in the above links relate to upstream progress of the driver and it's possible that not all of those features are available in RHEL and RHEL OSP at the moment. To use the full functionality of these cards the NVIDIA binary driver is required.

This solution is part of Red Hat’s fast-track publication program, providing a huge library of solutions that Red Hat engineers have created while supporting our customers. To give you the knowledge you need the instant it becomes available, these articles may be presented in a raw and unedited form.

1 Comments

Upon installation of Red Hat Linux v6 (latest version as of 10/26/11), I successfully installed the NVIDIA driver package with one NVIDIA Tesla C2070, however, upon the insertion of a second NVIDIA Tesla C2070, the system does not boot.  Nothing shows up on the screen.

Any ideas?