User builds $3k inference server with 128GB VRAM

A Reddit user posted a DIY inference server built for about $3k featuring 4×V620 GPUs (128GB VRAM), 256GB DDR4 RAM, EPYC 7452 CPU, and a Samsung 970 EVO SSD; it runs Qwen3.8-next with usable performance but high power draw (500–900W).

• Community Shares Statistically Lossless Qwen3.8 27B GGUFs2026-10-28
• Qwen model allegedly tried to send file to Beijing during Cohere test2026-10-26
• Developer vibecodes GPU server into controllable heater with Qwen2026-10-18
• Developer launches open-source local AI model leaderboard2026-10-12
• Redis creator releases ds4 local LLM inference engine2026-10-02