ZML is a production inference stack, purpose-built to decouple AI workloads from proprietary hardware.

Any model, many hardwares, one codebase, peak performance. Compiled directly to NVIDIA, AMD, TPU, Trainium for peak hardware performance on any accelerator. No rewriting.

• ZML @ GitHub.

Related contents:

• Introducing ZML/v2 @ ZML.

Very fast server provisioning for your data centre.

Self-service, remote installation of Windows, CentOS, ESXi and Ubuntu on real servers turns your data centre into a bare metal cloud.

• MAAS @ Launchpad Git.

Related content:

• Bare-Metal Forecasting : avec OpenBareMetal, c'est pas Versailles ici ! @ Open Source Experience's YouTube :fr:.

Simple way to manage your remote machines and SSH keys.

Bare metal servers are awesome. They let you pick where to run your software and how to deploy it. You get full control to make the most of the server's resources. No limits, no compromises. That's real freedom.

Viking makes it easier to work with them.

Bare metal to production ready in mins; your own fly server on your VPS.

Sidekick helps you along all the steps of deployment on your VPS. From basic setup to zero downtime deploys, we got you! ✊

Bare metal host provisioning for Kubernetes