Alibaba to offer Nvidia’s physical AI development tools in its AI platform (techcrunch.com)

Alibaba is integrating Nvidia’s Physical AI software stack into its Cloud Platform for AI, giving customers access to tools that build photorealistic 3D replicas of real environments and generate synthetic data for training models used in robotics, autonomous vehicles, and smart spaces like factories and warehouses. Financial terms weren’t disclosed, but the move pairs Nvidia’s end-to-end stack (designed to leverage its GPUs and software ecosystem) with Alibaba’s cloud and model-development infrastructure, enabling simulated data pipelines that can accelerate perception, planning, and agentic workflows without relying solely on costly real-world collection. The partnership is significant because synthetic, physics-aware datasets can dramatically speed iteration and improve safety for embodied AI — reducing annotation cost, increasing domain coverage, and enabling edge-case training for AVs and industrial robots. It also complements Alibaba’s broader AI push: the company is expanding data-center capacity globally, increasing AI investment beyond its prior $50B plan, and recently released Qwen 3-Max (claimed at 1 trillion parameters) for coding and agentic use. Together, these moves strengthen China-based cloud and simulation capabilities for large-scale model development and deployment, while highlighting Nvidia’s continued strategy of embedding its stack across major cloud providers worldwide.