GitHub - cubebecu/writeups: Field notes on hardware, local ML inference, and building things to understand them.

	Bringing an Aftermarket RTX 3090 Back to Life Bought a used RTX 3090 for local ML inference. It throttled on day one. Diagnosed, repasted, verified — hot spot dropped by 23°C.
+	Local LLM as a Coding Agent — A Practical Evaluation 🚧 Coming soon

Name		Name	Last commit message	Last commit date
Latest commit History 18 Commits
gpu-service		gpu-service
local-agentic-coding-evaluation		local-agentic-coding-evaluation
raw		raw
README.md		README.md

Provide feedback