On: How I learned to stop worrying and love userspace ...

[sweeter]

I'm always wondering how people are affording kubernetes clusters and all that to run stuff like ollama.

[skinkestek → sweeter]

The author is using https://k3s.io/ not the full k8s, so it doesn't have to be extremely expensive.

[tbrownaw → skinkestek]

K3s is a proper full k8s and has instructions for running multiple-node clusters, it just ships with batteries included and with defaults that play nice with having a single node.

[skinkestek → tbrownaw]

Thanks!

Today I learned!

[suprjami → sweeter]

You can run a local LLM on a $100 minipc with Vulkan GPU acceleration and get a usable token generation count.