NVIDIA Tesla Graphics Driver 348.07 64-bit

New in Release 348.07:

- Driver update for the R346 branch
- Adds CUDA 7.0 support

CUDA 7.0 Toolkit Release note:

- On POWER8 systems, users should add the /usr/lib/powerpc64le-linuxgnu/mesa path to the LD_LIBRARY_PATH environment variable in order to use the Mesa GL libraries. This information replaces this General CUDA Known Issue: "The POWER8 driver installation incorrectly overrides the Mesa GL alternative. To access the Mesa GL libraries, manually select the Mesa GL alternative by running the following command: sudo update-alternatives --config powerpc64le-linuxgnu_gl_conf."
- The default stream, used either when 0 is passed as a cudaStream_t or by APIs that operate on a stream implicitly, can now be configured to be a separate stream per thread that will not synchronize with other streams. Currently, operations in the default stream will serialize with other streams (see Stream Synchronization Behavior in the API documentation). This new behavior is not enabled by default and can be controlled per compilation unit with the --default-stream nvcc option.
- Added a method to the CUDA Driver API, cuDevicePrimaryCtxRetain(), that allows a program to create (or to access if it already exists) the same CUDA context for a GPU device as the one used by the CUDART (CUDA Runtime API) library. This context is referred to as the primary context, and this new method allows for sharing the primary context between CUDART and other threads, which can reduce the performance overhead of creating and maintaining multiple contexts per device.
- Unified the device enumeration for CUDA, NVML, and related tools. Variable CUDA_DEVICE_ORDER can have a value of FASTEST_FIRST (default) or PCI_BUS_ID.
- Instrumented NVML (NVIDA Management Library) and the CUDA driver to ignore GPUs that have been made inaccessible via cgroups (control groups). This enables schedulers that rely on cgroups to enforce device access restrictions for their jobs. Job schedulers wanting to use cgroups for device restriction need CUDA and NVML to handle those restrictions in a graceful way.
- Multi-Process Service (MPS) is now supported on GeForce products in addition to Tesla and Quadro products.
- Multi-Process Service now supports concurrent execution of GPU tasks on multiple GPUs at once. At startup, the MPS server attempts to spin up a context on each visible GPU and make it available to the client application.
- The Windows and Mac OS X installers are now also available as network installers. A network installer is much smaller than the traditional local installer and downloads only the components selected for installation.

Maximus Systems:

- For Maximus systems (Quadro + Tesla in the same system), download the latest recommended Quadro driver.
- Quadro drivers are qualified for workstations and rendering applications, including Maximus configurations that use Quadro for visualization and Tesla GPUs for compute acceleration.

Supported K-Series:

- Tesla K80
- Tesla K40c
- Tesla K40m
- Tesla K40s
- Tesla K40st
- Tesla K20Xm
- Tesla K20m
- Tesla K20s
- Tesla K20c
- Tesla K10
- Tesla K8

Supported C-Class:

- Tesla C2075
- Tesla C2070
- Tesla C2050
- Tesla C1060

Supported M-Class:

- M2090
- M2075
- M2070
- M2070-Q
- M2050
- M1060

Supported X-Class:

- Tesla X2070
- Tesla X2090