Dreamscale launches cloud inference for physical AI: 8+ robots per H100, 320ms DreamZero latency
arnie_hacker · x · 2026-10-16
Dreamscale Labs launched a real-time cloud inference platform for robot deployment: the UI reduces setup to "run <model> on <robot> to <do something>", no local GPU needed. The company claims end-to-end optimization of NVIDIA's DreamZero world-action model delivers 320 ms latency for the 14B model, 40–50% lower tail latency than existing cloud providers, and 8+ concurrent robots on a single H100 — beating NVIDIA's reported speed with half the silicon. It supports models from NVIDIA, Ai2, Physical Intelligence and others, plus user-supplied fine-tuned checkpoints.
More from Embodied
- WEDGE turns navigation agents' own failures into targeted training worlds — mohitban47 · 2026-10-16
- Knödel Watch: animated vocabulary cards on ESP32-S3, open-sourced 4 days after unboxing — alexcovo_eth · 2026-10-16
- SlipSense fuses pressure and vibration for 23ms zero-shot robotic slip detection — micoolcho · 2026-10-16
- Stanford's Fei-Fei Li team open-sources OpenWAM, a unified world-action model framework for robots — jiqizhixin · 2026-10-16
- Robot folding leaderboard: GPT-6 Astra leads at 1349 Elo, Muse Wonder trails at 1034 — qinzytech · 2026-10-16
- NVIDIA's TacitVLA: robots reason without reading their own thoughts, 3.4x lower latency — arankomatsuzaki · 2026-10-16