From Lamini, we have a private AMD GPU cluster, ready to serve any one who want to try MI300x or MI250 with inference and tuning.

We just onboarded a customer to move from openai API to on-prem solution, currently evaluating MI300x for inference.

Email me at my profile email.