Microchip’s new reference design combines PolarFire FPGA technology with CoaXPress connectivity to enable low-power, high-speed image capture for industrial edge vision systems.
Microchip has introduced its Edge Vision Capture and High-Speed Video Interface using CoaXPress Reference Design, providing developers with a ready-to-use platform for building high-performance edge vision applications across industrial, medical, robotics and automotive sectors.
The reference design is built around the PolarFire MPF100T FPGA, combining low power consumption with high-speed image processing capabilities. It supports the 12.5 Gbps CoaXPress (CXP) communication standard, enabling reliable, low-latency transmission of high-resolution video between cameras and host systems.
Designed to reduce development time, the platform integrates the essential hardware needed for machine vision applications. On the device side, the design features a PolarFire FPGA in the FCSG325 package, a 12.5 Gbps CoaXPress PHY, an Amphenol FCI connector, a Micro-BNC transmission connector and a JTAG programming interface. The host side includes a matching 12.5 Gbps CoaXPress PHY, an HPC FMC connector and a Micro-BNC receiver interface for seamless communication.
According to Microchip, the reference design delivers several advantages, including lower power consumption for energy-efficient operation, a cost-optimised footprint that reduces overall bill of materials (BOM) and PCB complexity, industrial-grade robustness for long-term deployments and seamless FPGA boot through SPI Flash for deterministic system startup.
The solution targets developers building edge AI cameras, automated inspection systems, medical imaging equipment, robotic vision platforms and intelligent transportation systems that require fast and reliable image acquisition.
By offering a validated hardware reference platform, Microchip enables engineers to accelerate product development while minimising design risks. The combination of PolarFire FPGA technology and the CoaXPress interface provides a scalable foundation for next-generation edge vision systems that demand high-speed data transfer, real-time image processing and reliable industrial performance.
Click here to review original reference design.
