Open seowalex opened 3 years ago
So... I haven't tried to set it up, but the generic answer I suspect is yes, since I'm seeing these things: https://medium.com/@adityathiruvengadam/cuda-docker-%EF%B8%8F-for-deep-learning-cab7c2be67f9 https://www.celantur.com/blog/run-cuda-in-docker-on-linux/
However, I will also point out the unpopular view that one shouldn't take it for granted that an ASV in the real world would have one or more on-board GPUs for processing, assuming that's what this question is about... That's not relevant for this simulated challenge.
More relevant to this challenge is that the server Docker image used to evaluate the solutions doesn't have CUDA set up, and there's no guarantee that the machines running the evaluation will have CUDA. So even if the generic answer is yes, I don't know if that will help you with VORC specifically.
Is there a way to make use of CUDA acceleration in our Docker images?