PhyAI lifts H20 inference throughput from 10 Hz to 36 Hz for Physical AI workloads
CyberRobooo · x · 2026-07-24
PhyAI raises MiniCPM-Robot throughput on H20 from 10 Hz to 36 Hz
This reply adds that the release is backed by PhyAI, a Physical AI inference framework aimed at fast edge inference and large-scale cloud RL rollouts.
- Designed for broader Physical AI workloads, not just MiniCPM-Robot
- Uses CUDA Graph to lift H20 throughput from 10 Hz to 33 Hz
- Custom Triton fused kernels push it further to 36 Hz
- The point is that open robotics needs not only model weights, but also deployment and inference tooling
Related event: OpenBMB Open-Sources MiniCPM-Robot for Offline Embodied AI(10 posts)→
More from Embodied
- NVIDIA open-sources FLORA, a robotics data engine ~125x cheaper than GPT-6 Astra — nvidia · 2026-10-27
- Embodied AI startups need a different operational discipline, investor says — ___Mufasaa · 2026-10-27
- π0.5 study: raising demos from 10 to 438 lifts Place Ball in Bowl success from 7% to 83% — stingning · 2026-10-27
- GPT-6 Astra controls dexterous hands zero-shot at just 5.2% success in 500 trials — stingning · 2026-10-27
- HandsOn benchmark: 66 verified tasks and 18,000 human demos for dexterous manipulation — stingning · 2026-10-27
- DexPort: a handheld interface with a real robot hand for dexterous-hand data collection — pabbeel · 2026-10-27