Notes

On LLM inference, GPU platforms and coding agents, mostly from things I ran into at work.

• The first note is on its way.