Selecting local coding models for high RAM, limited VRAM setups
Frail_Waif · reddit · 2026-08-30
A user seeks advice on running local agentic coding models on a workstation with 256GB RAM but limited GPU power (RTX A4500 20GB + 940MX 12GB). The debate is between using a quantized Qwen 32B model that fits the VRAM versus a larger model utilizing the massive RAM. The user plans to use WSL2 and llama.cpp, asking if the hardware is sufficient for productive 8-hour workdays to convince colleagues.
More from coding & agent
- Animated teardown of XBOW's Chrome full-chain exploit that won Google's first $250K VRP bounty — moyix · 2026-10-15
- pstack's durable core: 10-20 line skills that distill agent best practices into principles — HankYeomans · 2026-10-15
- 'Children of Elon' forum lets AI agents post and teach each other while humans watch — AiJohnAllen · 2026-10-15
- VoidZero open-sources Void: SDK, CLI and platform to deploy Vite apps on Cloudflare — cnakazawa · 2026-10-15
- Running 2 agents on the same data breaks everything — Omnigraph offers a fix — sarahdrinkwater · 2026-10-15
- Evan You presents 'Developer Tooling in the Agentic Era' at ViteConf 2026 — cnakazawa · 2026-10-15