Based on the Synaptics Torq Edge AI platform, it delivers transformer-class inference directly into embedded hardware, supporting modern AI models without the need for proprietary tooling platforms. A modular architecture with configurable compute modules, motherboards, and expansion cards simplifies the development of real-world vision systems.
The kit incorporates the Synaptics SL2619 processor and the Synaptics Torq Edge AI platform, based on an open software stack built on Yocto Linux and open technologies such as IREE and MLIR, avoiding dependence on proprietary tools and facilitating the portability of modern AI models. It also integrates the T1 NPU along with Google's Coral Open NPU technology to run computer vision models and multimodal applications directly on the device.
