This page was machine-translated and may differ from the original. View original
Aiming to develop multimodal AI prototypes based on Coral Open NPU
Mouser announced on the 9th that it is supplying the development kit through its distribution network. The SL2600 is an edge AI development kit designed to run multimodal and transformer-class inference on embedded IoT hardware.
This kit is based on Synaptics' AI-native Torq edge AI platform. Torq is configured to run AI models, such as Vision Transformers and multimodal pipelines, by utilizing Google's RISC-V-based Coral Open NPU.
The software environment is centered around Yocto Linux and IREE·MLIR-based compilers and runtimes. Through this, Synaptics presented a structure that reduces dependence on proprietary toolchains or specific models.
The SL2600 features a modular hardware architecture consisting of a computing module, an I/O baseboard, and an expandable daughter card. Developers can configure vision system prototypes by utilizing integrated connectivity, programmable I/O, and interfaces tailored to real-world product environments.
Daughter card interfaces include MIPI DSI, MIPI CSI-2, a 40-pin JTAG header, and a 4-pin PoE+ connector. This configuration is applicable to the development of various edge AI systems, such as camera-based vision processing, industrial equipment, robots, and smart devices.
This supply aligns with the trend of expanding AI inference capabilities from a cloud-centric architecture to field terminals. As demand for real-time processing, reduced latency, and alleviated data transmission burdens increases in industrial equipment and smart devices, the scope of application for edge AI development platforms is also expanding.
More information about the Astra Machina SL2600 development kit can be found on the Mouser product page.
본 기사에 대한 정정·반론·추후보도 청구는 보도 청구 안내를, 그간 게재된 보도문은 정정·반론보도 모아보기를 참고해 주세요.















