Infinigence AI Open-Sources APXInf for Embodied Edge Inference on Jetson Thor

Infinigence AI, alongside Tsinghua and Shanghai Jiao Tong Universities, has open-sourced APXInf, an edge inference engine reducing latency on NVIDIA Jetson Thor from ~278 ms to ~26 ms, enabling real-time embodied AI control. The release supports Jetson Orin, Thor, RTX 4090, and plans domestic-chip and robot OS integration.

This Wire brief sits within Fusion42's coverage of AI & ML.