search

NVIDIA / gpu-operator

NVIDIA GPU Operator creates, configures, and manages GPUs in Kubernetes
https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/index.html
Apache License 2.0
1.82k stars 297 forks source link

Nvidia Operator Documentation for installing on Amazon Linux 2023 and Bottle Rocket AMIs #946

Open dgr237 opened 2 months ago

dgr237 commented 2 months ago

1. Quick Debug Information

2. Feature description

I am interested in understanding whether the Nvidia operator is compatible with Bottle Rocket and/or Amazon Linux 2023. We are currently building our own custom AMIs based on the EKS Optimized AMIs and installing specifc Nvidia Drivers which have been certified by the business.

We are looking at the Nvidia Operator as an alternative to building custom AMIs and leveraging the Nvidia Operator. This would greatly reduce the complexity around building a custom AMI and releasing to the business for them to test and validate and iterating over this process several times to get a version of the Nvidia drivers which correctly install on the AMI and which meets the needs of the business.

We would therefore be interested in the documentation for the Nvidia Operator to detail compatibility with Amazon Linux 2023 and/ or Bottle Rocket along with detailed steps on how to install the operator and custom driver versions.

cdesiniotis commented 3 weeks ago

Hi @dgr237, we do not currently support Bottle Rocket or Amazon Linux 2023. The operator could technically work on Amazon Linux 2023 as long as the NVIDIA drivers are preinstalled. Although I have not tried this out myself.

Support for Amazon Linux 2023 is in our roadmap.