Running DeepSeek V4 Flash Locally on 2x DGX Sparks Delivers Prosumer-Grade Performance
andrewchen · x · 2026-08-13
Andrew Chen shared his experience running DeepSeek V4 Flash (0731) locally. On a setup equipped with 2x DGX Sparks, the model performs exceptionally well and very fast.
- Excellent Performance: Features low Time To First Token (TTFT) and maintains a highly usable generation speed of around 50 tok/s.
- Best Setup: He regards this as the best prosumer-grade local AI hardware setup right now.
More from Infra
- New hardware day: local LLM deployment the moment it arrives — lxfater · 2026-10-16
- Together AI's score centering fixes RL training collapse under quantized inference — nrehiew_ · 2026-10-16
- Inference is splitting prefill and decode across specialized chips, not just H100s — AccBalanced · 2026-10-16
- Rebasely launches in beta: Git hosting built for humans and coding agents — handotdev · 2026-10-16
- Drama3 TTS lands in audio.cpp: up to 5x real-time on RTX 5090, <8GB VRAM mode too — Acceptable-Cycle4645 · 2026-10-16
- Three CoreWeave co-founders have sold $3.3B in shares since IPO amid credit worries — SumitGup · 2026-10-16