Microchip releases smaller PolarFire FPGA Ethernet Sensor Bridge Rev 2.0
Microchip Technology has released Revision 2.0 of its PolarFire FPGA Ethernet Sensor Bridge, a production-ready hardware platform designed to standardise sensor integration for edge AI systems. The updated board is 60 per cent smaller than its predecessor, supports up to four cameras, and is powered via USB-C, a change the company says simplifies rack deployment. Microchip says it is also being offered at a lower price point than the first generation, though it did not disclose specific pricing in its announcement.
The board is built to work with NVIDIA's Holoscan Sensor Bridge technology and is compatible with NVIDIA Jetson and IGX edge computing platforms, targeting AI-driven workloads in medical imaging, industrial automation and humanoid robotics. On-board optical latency measurement circuitry enables end-to-end latency validation from sensor capture through to AI inference using NVIDIA's Latency Display Analysis Tool.
Technical specification
At the core of the design is Microchip's PolarFire FPGA, which the company positions as offering low power consumption alongside built-in security and safety features suited to long-running edge deployments. The board supports a range of sensor and video interfaces, including MIPI CSI-2, SLVS-EC 2.0, 12G-SDI, HDMI and DisplayPort, through an FPGA Mezzanine Card expansion connector. A PMOD connector adds further extensibility without requiring a redesign of the base platform.
Integrated power management is handled by Microchip's own MCP16701 PMIC, validated for PolarFire FPGAs. The package ships with cameras, cables, a reference design, board schematics and RTL development support via Microchip's Libero SoC Design Suite, alongside integration with the NVIDIA Holoscan SDK.
Shakeel Peera, vice president of Microchip's FPGA business unit, said the aim was to free developers from the burden of proprietary sensor interface integration. "Developers want to spend their time building high-value edge AI applications, not stitching together proprietary sensor interfaces," he said.
Market context
The release lands as edge AI hardware becomes an increasingly crowded market. FPGA vendors including Intel (Altera) and Xilinx (now part of AMD) compete directly with Microchip's PolarFire line for edge inference workloads, while dedicated AI accelerator startups are also targeting the industrial and robotics segments. The shift to 10Gb Ethernet as the backbone for multi-sensor architectures is a broader industry trend, driven by the need to standardise data pipelines across heterogeneous sensor types without custom silicon.
Microchip's partnership with NVIDIA is significant context here. NVIDIA's Holoscan platform has been gaining traction in the medical AI and robotics space, and hardware vendors that offer validated, production-ready integrations with Holoscan gain a route to market through NVIDIA's developer ecosystem. The Rev 2.0 board's NVIDIA Jetson compatibility also keeps it relevant as Jetson remains the dominant platform for on-device AI inference in industrial and robotic applications.
Regulatory and standards read-across
Edge AI deployments in medical and industrial settings face growing regulatory scrutiny. In the EU, the AI Act's high-risk classification covers AI systems used in medical devices and safety-critical industrial machinery, imposing requirements around transparency, data governance and human oversight. Hardware vendors supplying into those verticals increasingly need to demonstrate that their platforms can support the auditability and reliability standards that compliance demands. Microchip's emphasis on built-in security features and deterministic latency will be relevant selling points for customers navigating those obligations.
The Rev 2.0 PolarFire FPGA Ethernet Sensor Bridge is available to purchase directly from Microchip or through its authorised distributor network. No volume pricing tiers or lead times were disclosed.