experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Include NCCL in the list of requirements for TensorFlow with GPU support

Browse Source 
PiperOrigin-RevId: 204409167
This commit is contained in:
Smit Hinsu 2018-07-12 19:45:23 -07:00 committed by TensorFlower Gardener
parent b4654e6c93
commit 59bfc71a71
1 changed files with 2 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tensorflow/docs_src/install/install_linux.md
View File

@ -511,6 +511,8 @@ on your system:
list of supported GPU cards. list of supported GPU cards.
* [GPU drivers](http://nvidia.com/drivers) that support your version of the CUDA * [GPU drivers](http://nvidia.com/drivers) that support your version of the CUDA
Toolkit. Toolkit.
* NCCL 2.2 to use TensorFlow with multiple GPUs. For details, see [NVIDIA's
documentation](https://developer.nvidia.com/nccl).
* The `libcupti-dev` library is the NVIDIA CUDA Profile Tools Interface. This * The `libcupti-dev` library is the NVIDIA CUDA Profile Tools Interface. This
library provides advanced profiling support. To install this library, library provides advanced profiling support. To install this library,
use the following command for CUDA Toolkit >= 8.0: use the following command for CUDA Toolkit >= 8.0: