The simplest micro benchmark for cupy CUDA containerized in NVidia AI Workbench

NVidia AI Workbench runs inside a containerized environment and I wanted an environment check that verifies the container has access and that docker/podman, the NVidia driver and Workbench are all on compatible versions.  

Containerization has no effect on performance. The CUDA code is pretty much a direct pass-through to the card.

Environment

  • NVidia AI Workbench
  • One local GPU
  • Windows 11 
  • Docker Desktop
  • Program running in a container built by the Workbench based on the PyTorch/Cuda image

Program

I access the containerized environment via a Jupyter Notebook visible to the browser on the Windows machine. This is a snapshot of the Jupyter Notebook.


It found a problem

The program demonstrated that there was a container adapter (or something) mismatch that recently happened.  cupy returned that it had access to the GPU but it turns out that it really didn't and that a Docker Desktop upgrade was needed to fix something driver-related.  The error message was

"CUDARuntimeError: cudaErrorSymbolNotFound: named symbol not found"

Revision History

Created 2024/07


Comments

Popular posts from this blog

Installing the RNDIS driver on Windows 11 to use USB Raspberry Pi as network attached

Understanding your WSL2 RAM and swap - Changing the default 50%-25%

Almost PaaS Document Parsing with Tika and AWS Elastic Beanstalk