Skip to content

vsi_gpu_nvidia_drivers.md

Latest commit

 

History

History
29 lines (22 loc) · 1.89 KB

vsi_gpu_nvidia_drivers.md

File metadata and controls

29 lines (22 loc) · 1.89 KB
copyright
years
2018, 2024
lastupdated 2024-07-24
subcollection virtual-servers

{{site.data.keyword.attribute-definition-list}}

Installing GPU drivers and software packages

{: #installing-gpu-drivers-and-software-packages}

You need to install the following software before you can use a GPU profile with your virtual server.

  • NVIDIA drivers{: external} - allows your operating system to communicate with the GPU.
  • CUDA Toolkit{: external} - development environment for high-performance GPU-accelerated applications.
  • cuDNN{: external} - (CUDA Deep Neural Network) library that is used for neural network applications.
  • NCCL{: external} - cluster communications library for multi-GPU and or multi-system communications.
  • BLAS (Basic Linear Algebra Subprograms) - routines that provide the building blocks for performing basic linear algebraic operations such as one of the following libraries:

If you are installing a machine learning framework, you need to install framework software such as one of these platforms in addition to the preceding software packages.

  • Caffe{: external} (Convolutional Architecture for Fast Feature embedding) - licensed, open source deep learning framework.
  • Tensorflow (Tensorflow 1.2+){: external} - provides an open source software library for numerical computation.