Designing smarter edge AI systems starts with PolarFire® FPGAs. Get low power, real-time responsiveness and flexible deployment whether you are running inference directly on the FPGA or streaming data to the NVIDIA® Holoscan platform and other accelerators.
With VectorBlox SDK 2.0, you can deploy neural networks directly on PolarFire FPGAs; no FPGA expertise is required. Achieve real-time video intelligence at the edge with unmatched power efficiency.
When robotics and autonomous systems demand high-performance AI, PolarFire FPGAs serve as the front end. The Ethernet Sensor Bridge aggregates and streams high-bandwidth sensor data to NVIDIA Holoscan or other accelerators to enable real-time robotic vision and multi-sensor fusion.
This solution streams multiple high-resolution video feeds into the NVIDIA Holoscan platform for low-latency AI inference in surgical, industrial and autonomous robots.
LVDS and JESD are commonly used protocols for image acquisition in ultrasound, MRI and CT/PET scanners. Designers of endoscopes are rapidly adopting SLVS-EC™ image sensors while using CoaXPress to transfer uncompressed video to the receiver.
PolarFire SoC FPGAs deliver up to 50% lower power than equivalent SRAM SoC FPGAs. We built PolarFire SoCs upon the award-winning, nonvolatile PolarFire FPGA platform. They feature a five-core Linux®-capable processor subsystem based on the RISC-V® ISA.
Our award-winning PolarFire FPGAs deliver the industry’s lowest power at mid-range densities with exceptional security and reliability. This family of products spans from 100K Logic Elements (LEs) to 500K LEs, features 12.7G transceivers and offers up to 50% lower power than competing mid-range FPGAs