Pycuda Docker environment

Image details

• Pycuda 2019.1.1
• CUDA 10.0 devel
• Python 3.6.8
• Ubuntu 18.04 Bionic

Required

• Follow nvidia-docker instructions

Option 1. Build Docker Image

• From within the directory of the Dockerfile execute the following command to build the image

docker build -t pycuda .

• To run with bash and a shared volume

docker run --runtime=nvidia -v <host_path>:<container_path> -ti pycuda:latest bash

Option 2. Pull image from Docker Hub

• Go to the repository page bryankp/pycuda
• Pull the latest image

docker pull bryankp/pycuda:latest

• To run with bash and a shared volume

docker run --runtime=nvidia -v <host_path>:<container_path> -ti bryankp/pycuda:latest bash

Test if pycuda is working correctly

• See the test_script README

Further Details

• Pycuda project page
• For CUDA troubleshooting issues see nvidia-docker instructions