A full stack platform for Edge AI from Google

Google launched Coral, a full-stack platform for edge AI that combines an AI-first hardware architecture with a unified developer experience to enable efficient, local AI processing. The platform provides software developers with tools to deploy PyTorch, JAX, or LiteRT models on any device using standards-based MLIR compiler toolchains, while hardware developers can leverage open-source, RISC-V based reference designs for custom silicon. This initiative aims to power the next generation of edge devices by bringing high-performance, ultra-low power AI to local environments.

A full stack platform for Edge AI https://developers.google.com/coral/guides/intro By combining an AI-first, standards-based hardware architecture with a unified developer experience, Coral enables efficient and local AI experiences at the edge. For software developers https://developers.google.com/coral/guides/software/get-started Deploy intelligent models at the edge. Seamlessly bring your PyTorch, JAX, or LiteRT models to any device using our standards-based MLIR compiler toolchains and simulators. For hardware developers https://developers.google.com/coral/guides/hardware/get-started Power the next generation of edge devices. Leverage flexible reference designs and our open-source, RISC-V based architecture to bring high-performance, ultra-low power AI to your custom silicon.