MathWorks has announced the release of a hardware support package for the Qualcomm Hexagon Neural Processing Unit (NPU), integrated within the Snapdragon processor range. This new support package facilitates the automated generation of code from MATLAB and Simulink models, specifically optimised for Qualcomm’s Hexagon NPU architecture.
It is designed to enhance data accuracy, ensure compliance with standards, and streamline the developer workflow. Using MathWorks software and Model-Based Design allows engineers to deploy high-quality C code for embedded applications, without requiring detailed knowledge of the NPU itself.
“Qualcomm Technologies’ mission is to empower engineers to create innovative products and applications that leverage the full potential of Snapdragon processors,” said Vinesh Sukumar, Senior Director of Product Management, Qualcomm Technologies. “As the demand for high-performance NPU applications continues to grow, the collaboration between MathWorks and Qualcomm Technologies enables developers to efficiently develop embedded applications optimized for the Hexagon NPU.”
The MathWorks hardware support package for the Hexagon NPU addresses several key challenges in the development process, including compilation issues, reliance on hand-crafted frameworks, manual numeric verification, and the difficulty of tracing code back to its model. Engineers without extensive knowledge of the Hexagon NPU or its toolchain can use the package to perform Processor-In-the-Loop (PIL) simulations directly in Simulink, using the Hexagon simulator. This ensures the optimised target code functions correctly, efficiently, and reliably before it is deployed on physical hardware.
“By harnessing the power of Model-Based Design and the advanced capabilities of the Hexagon NPU, we are making it easier for engineers to create high-quality NPU applications,” said Jim Tung, MathWorks Fellow. “Our collaboration with Qualcomm Technologies is focused on removing manual processes in favour of highly automated workflows to help engineering organizations reduce time to market and ensure optimal performance.”
The automatic code generation features of MATLAB and Simulink are trusted globally by industries such as automotive, aerospace, medical devices, and more. MathWorks software already supports code optimisation for a variety of processors, including the ARM Cortex-M, ARM Cortex-A, ARM Cortex-R, Infineon AURIX TC3x and TC4x, as well as GPUs from ARM and NVIDIA.
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.
