Microchip Technology Inc. today announced the availability of its PolarFire FPGA Ethernet Sensor Bridge that works with the NVIDIA Holoscan sensor processing platform to help developers build artificial intelligence (AI)-based sensor processing systems.

PolarFire FPGAs are the first solution in Microchip’s platform to enable multi-protocol support, compatible with MIPI CSI-2-based sensors and MIPI D-PHY physical layers.

Future solutions will support a wide range of sensors with various interfaces, including △SLVS-EC™ 2.0 △12G SDI △CoaXPress® 2.0 and △JESD204B.

With this platform, developers will now be able to leverage the power-efficient technology of PolarFire FPGAs with low-latency communications and multi-protocol sensor support, while leveraging the power of the NVIDIA Holoscan ecosystem.

The NVIDIA Holoscan platform simplifies the development and deployment of AI and high-performance computing (HPC) applications at the edge to deliver real-time insights. It also integrates the essential hardware and software systems required for low-latency sensor streaming and network connectivity into a single platform.

The platform includes optimized libraries to help process data quickly, sample AI models to quickly jump-start AI inference pipeline development, template applications to support rapid prototyping, and core microservices to perform a variety of tasks such as streaming and video processing.

The PolarFire FPGA Ethernet Sensor Bridge can connect real-time sensor data to NVIDIA Holoscan, NVIDIA IGX, and NVIDIA Jetson platforms to support edge AI and robotics. This enables new edge-to-cloud applications, enables AI/ML inference, and accelerates AI adoption across a wide range of markets, including healthcare, industrial, and automotive.

The NVIDIA Holoscan sensor bridge efficiently manages high-bandwidth data transmitted over Ethernet from a variety of sensors using Microchip’s low-power PolarFire FPGA technology. It enables real-time, high-performance edge AI processing on NVIDIA AI platforms. This power-efficient design is also ideal for applications where space is limited and energy or cost efficiency is critical.

PolarFire FPGAs provide embedded security and safety features to address security concerns that may arise in sensor applications, protecting against potential cyber threats and ensuring the physical device, design, and data integrity of the system.

It also features Single Event Upsets (SEU) immunity, providing high reliability even in environments exposed to radiation, such as space, high-altitude applications, and medical environments. SEU tolerance helps reduce data corruption and system failures.

“The Ethernet Sensor Bridge is based on Microchip’s power-efficient, secure and reliable PolarFire FPGA platform,” said Bruce Weyer, vice president of Microchip’s FPGA business unit. “Combining our flexible FPGA architecture with NVIDIA’s advanced AI platform and multi-protocol support enables developers to develop real-time solutions for powerful AI-based edge applications.”