Building a $15k Local Inference Rig for DeepSeek V4 Flash: MI210 vs A40?
_TheWolfOfWalmart_ · reddit · 2026-09-12
A Reddit user seeks GPU advice for a $15k work rig to run DeepSeek V4 Flash and similar models locally with at least 128GB VRAM and minimal quantization, fitting 3 cards in a Dell R740. Candidates: 3× AMD MI210 (192GB) vs 3× NVIDIA A40 (144GB); they're asking for real token-gen and prompt-processing benchmarks and say cloud rental for testing is unavailable.
More from Infra
- SemiAnalysis: OpenAI reportedly negotiating ~1GW TPU commitment while seeking sites — dylan522p · 2026-10-28
- I tested 11 video upscaling methods for AI video: local, Topaz and cloud compared — lopas99 · 2026-10-28
- Thermo-backed CTO says AI is in its dial-up phase, touts disaggregated compute for an ultrafast 'broadband moment' — beffjezos · 2026-10-28
- DeepSeek-V4.1-Flash Q2 runs at 41 tok/s on one RTX 3090, 90% token match vs API — musicurgy · 2026-10-28
- USearch v3 RC ships after 1,000 days, its libraries now embedded in 550K-star projects — viglovikov · 2026-10-28
- SK Hynix says memory demand will outpace supply through 2030 — Chemical_Hawk_6307 · 2026-10-28