Welcome
CosmicAC
Self-hosted GPU compute for machine learning and compute-intensive workloads. Deploy CosmicAC on your own Kubernetes cluster to run GPU Container jobs and serve open-source language models with Managed Inference, all on hardware you own and control.
What CosmicAC offers
- Dedicated GPU containers: A GPU Container job is an isolated KubeVirt virtual machine (VM) with dedicated GPU hardware and hardware-level isolation, accessible through an interactive shell session.
- Language model serving: Managed Inference runs open-source models on your cluster without you building a serving stack from scratch.
- Runs on your infrastructure: CosmicAC is deployed onto your own Kubernetes cluster, so your data and workloads never leave your hardware.
- Full environment control: GPU Container jobs give you full control over the VM environment, including the operating system, packages, and scripts.