topic Re: Does Databricks Container Services (DCS) support for GPU containers with Databricks Runtime 11.3 in Machine Learning

Does Databricks Container Services (DCS) support for GPU containers with Databricks Runtime 11.3 LTS and higher?

ppang — Sat, 20 May 2023 22:23:29 GMT

I have been trying to start a cluster using DCS with GPU containers (https://github.com/databricks/containers/tree/master/ubuntu/gpu), but was only successful with Databricks Runtime 10.4 LTS and lower.

With Databricks Runtime 11.3 LTS and higher, I got the error:

"Internal error message: Spark error: Driver down cause: driver state change"

Does DCS supports GPU containers for 11.3 LTS and higher?

Re: Does Databricks Container Services (DCS) support for GPU containers with Databricks Runtime 11.3

jessysantos — Mon, 27 May 2024 22:33:26 GMT

Hello @ppang !

Since you posted your question, the repository you shared has received an update, which includes the following warning:

"Using conda in DCS images is no longer supported starting Databricks Runtime 9.0. We highly recommend users to extend cuda-11.8 examples. We no longer support cuda-10.1 and cuda-11.0 compatibility with latest databricks runtime."

It's likely that the issue you encountered was related to a CUDA incompatibility.

Best regards,

Jéssica Santos