Managed Kubernetes
Container platforms tuned for AI. 32 providers on EveryNeocloud so far.
Kubernetes is open-source software, first released by Google in 2014, that runs programs in containers across many servers and restarts them when they fail. Most AI clouds now offer it as a managed service tuned for GPUs, so customers can run training and inference without managing the machines underneath. CoreWeave built its cloud around Kubernetes from the start.
Why AI teams use it
Containers package a model with all the software it needs, so it runs the same way on any machine. Kubernetes spreads those containers across GPU servers, scales them up when traffic rises and down when it falls, and moves work off broken servers. Add-ons let it schedule GPU jobs in groups, so a training run starts only when all its GPUs are free.
Who offers it
All three big clouds have managed Kubernetes services. Among neoclouds, CoreWeave runs its own Kubernetes service on bare metal, and Nebius, Lambda and Nscale offer managed clusters. Developer clouds such as DigitalOcean and Linode include it, and Rackspace Spot sells Kubernetes clusters at auction prices. Many research teams still prefer Slurm, the older job scheduler of supercomputers.
Famous managed Kubernetes providers
Still to be written: Google Cloud Platform.