Dual rtx 3090 build

Posted by Sufficient_Phone_242@reddit | LocalLLaMA | View on Reddit | 53 comments

Dual rtx 3090 build

Joining this community sparked a new hobby and interest in software engineering that I had lost.

So I made this dual rtx 3090 build mostly for inference , I know I won’t be replacing chatgpt anytime soon but what tool stack would help it be usable in a work environment ? Must MCP servers or custom tools/scripts ?

Currently using VScode preview with qwen3.6 27b and an nginx server, Im mostly interested in agentic work with usable context or at least a better knowledge of code base ( RAG pipeline?)

Been already such a helpful community , hopefully local llms continue to grow because I fear cloud will become unaffordable at a consumer level