Picked up a $1,100 RTX 3090 FTW3 for a headless local LLM server — seeking configs
NighthawkXL · reddit · 2026-11-01
The author bought a low-mileage EVGA RTX 3090 FTW3 (24GB) for $1,100 and plans to put it in a headless Ubuntu Server box (Ryzen 5 5600X, 64GB RAM) as a LAN API backend for assistant and coding inference, replacing a cramped RTX 4070 12GB setup. They ask the community for:
- Coding models for 24GB suited to reverse engineering DOS/16-bit Windows titles (x86 assembly, Ghidra/IDA pseudo-code) at 32k-128k context, 20-22GB weights with KV cache headroom.
- Headless serving and on-demand model swapping backends that won't starve co-hosted services like Jellyfin and AdGuard, preferably containerized.
- FTW3 thermals on Linux: power limits via nvidia-smi -pl, headless undervolting, and fan-curve setups (coolbits, nvidia-settings, systemd services) for the card's notoriously hot rear GDDR6X.
More from Infra
- Depth-growing 3B models beat dense training on GSM8K (19% vs 12%) while saving 26% pretraining compute — burny_tech · 2026-11-01
- From-scratch CUDA engine runs full GLM-5.3 on 3-6 DGX Sparks; steering vectors flip output language — burny_tech · 2026-11-01
- How NVIDIA's instruction scheduling model works: static latencies and scoreboards — mike64_t · 2026-11-01
- Sequoia's Shaun Maguire builds SpaceX model projecting $731B revenue by 2028 — FinanceYF5 · 2026-11-01
- SK hynix sold out through 2029: AI selloff is about cash flow, not demand, argues Puru Saxena — danielrock · 2026-11-01
- Gemma 4 reportedly running in a space data center under Google's Project Suncatcher — Saboo_Shubham_ · 2026-11-01