What it does
dstack is a control plane for provisioning and orchestrating GPU compute. It works with GPU clouds, Kubernetes, VMs and on-prem clusters, and covers development, training and inference. It supports NVIDIA, AMD, Google TPU and Tenstorrent accelerators.
Use cases
- 01Run model training across several GPU clouds
- 02Serve inference on a Kubernetes cluster
- 03Manage GPU capacity on bare-metal servers