Public | Automated Build

Last pushed: 2 years ago
Short Description
F@H (Folding @Home) image built to use Nvidia CUDA.
Full Description

docker-cuda-fah

Folding @Home image with CUDA support.

Usage:

On a host with a CUDA device and drivers.

    docker run \
      --device=/dev/nvidia0:/dev/nvidia0 \
      --device=/dev/nvidiactl:/dev/nvidiactl \
      --device=/dev/nvidia-uvm:/dev/nvidia-uvm \
      -it \
      --rm cuda-fah \
      <FAHClient command line options>

Or, as the base of another image.

    FROM ozzyjohnson/cuda-fah

Next:

  • EXPOSE ports for the control program.
  • We can probably just drop a config file in the directory mounted to /data and run without options. Have to test.
  • Complete testing, fah.stanford.edu is unreachable as I commit this.
Docker Pull Command
Owner
ozzyjohnson
Source Repository