Página inicial Explorar Ajuda
Registrar Entrar
jeremy
/
HON95-wiki
mirror de https://github.com/HON95/wiki
1
0
Fork 0
Arquivos Issues 0 Wiki
Tree: f5a0a18352
Branches Tags
HON95-wiki
/
se
/
frameworks
/
cuda.md

cuda.md 2.1 KB
Histórico Raw

title: CUDA breadcrumbs:

  • title: Software Engineering
  • title: Frameworks --- {% include header.md %}

NVIDIA CUDA (Compute Unified Device Architecture) Toolkit, for programming CUDA-capable GPUs.

Resources

Installation

Linux

The toolkit on Linux can be installed in different ways:

  • Through an an existing package in your distro's repos (simplest and most compatible, but may be outdated).
  • Through a downloaded package manager package (up to date but may be incompatible with your installed NVIDIA driver).
  • Through a runfile (same as previous but more cross-distro and harder to manage).

Note that the toolkit requires a matching NVIDIA driver to be installed.

Ubuntu (package from main repos):

  1. Update your NVIDIA driver.
    • Typically through the "Driver Manager" on Ubuntu-based distros, which installs it through the package manager.
    • Check which version you have installed with dpkg -l | grep nvidia-driver.
  2. Install the CUDA toolkit: apt install nvidia-cuda-toolkit

Ubuntu (downloaded package or runfile):

See NVIDIA CUDA Installation Guide for Linux (NVIDIA).

Running

  • Gathering system/GPU information with nvidia-smi:
    • Show overview: nvidia-smi
    • Show topology matrix: nvidia-smi topo --matrix
    • Show topology info: nvidia-smi topo <option>
    • Show NVLink info: nvidia-smi nvlink --status -i 0 (for GPU #0)
    • Monitor device stats: nvidia-smi dmon
  • To specify which devices are available to the CUDA application and in which order, set the CUDA_VISIBLE_DEVICES env var to a comma-separated list of device IDs.

{% include footer.md %}