How can i install cuda on ubuntu 16.04 update#

The command lspci is used to display the devices connected to PCI buses. Verify the GPU is CUDA capable $ lspci | grep -i nvidia

You can refer to this article for NVIDIA graphics driver installation and functionality checking.ģ. Verify that NVIDIA graphics driver has been installed properly.

How can i install cuda on ubuntu 16.04 update#

Update apt package index and install the newest version of all currently installed packages $ sudo apt-get updateĢ. The pre-installation steps are primarily checking if the system environment is ready for Cuda toolkit installation.ġ. The driver installation is not covered in the post. Prior to the installation, ensure that you have installed NVIDIA driver for Ubuntu. You can refer to the article if you want to go straight to CUDA 9.2 instead of 9.1. Update: I have written a follow-up post about installation of CUDA 9.2.

The official installation guide is available at the NVIDIA website and can be referenced when following the steps outlined in this post. If you are using Ubuntu 16.04 (Xenial) and want to install the recent release of CUDA toolkit (version 9.1), this post may help. To harness the GPU power, NVIDIA develops and provides CUDA toolkit that can be used as the development environment and libraries for GPU-accelerated applications. NVIDIA graphic cards have gained popularity among machine learning researchers and practitioners as the base hardware for GPU computing. It is not uncommon to perform computation for linear algebra, image and video processing, machine learning (especially deep learning), graph analytics, and so forth on GPU. With advances in GPU technologies, performing complex computation is not an exclusive feat of multicore CPUs anymore.