DeepSeek V4.1 Flash hits 103 tok/s on two-box lucebox cluster, beating Mac Studio on price
QuixiAI · x · 2026-10-29
QuixiAI reports running DeepSeek V4.1 Flash at 103 tok/s on a cluster of two lucebox boxes, claiming it is cheaper and faster than a Mac Studio setup. The lucebox team says tensor parallelism is being optimized to near-DGX efficiency with low latency as the key, while admitting networking still lags DGX for models that don't fit in a single box.
More from Infra
- Google Cloud launches Rubin-powered A5X instances scaling to 80,000 GPUs per site — Beth_Kindig · 2026-10-30
- Sam Altman calls widespread local inference "quite terrible" for global DRAM allocation — Hesamation · 2026-10-30
- Microsoft's Maia 300: 2x chips per rack, 2.5x memory bandwidth, up to 7x throughput from 2027 — Beth_Kindig · 2026-10-30
- Evo 2 genomic model now runs on vLLM: 7B fits 1-megabase context on a single H100 — anshulkundaje · 2026-10-30
- Meta's GenRecs Tokenization lifts ad conversions +0.35% with semantic vocabulary — Meta_Engineers · 2026-10-30
- Parallel Language Models paper: multi-token generation at AR quality, up to 5.7x speedup — aaron_lou · 2026-10-30