Low-power FPGA accelerates embedded vision
The company now extends its high-resolution smart embedded vision FPGA offerings with enhanced high-speed imaging interfaces, an intellectual property (IP) bundle for image processing and an expanded partner ecosystem. To further address design requirements for intelligent vision systems, Microchip has added a Serial Digital Interface (SDI) IP to transport uncompressed video data streams over coaxial cabling. The interface comes in multiple speeds: HD-SDI (1.485 Gbps, 720p, 1080i), 3G-SDI (2.970 Gbps, 1080p60), 6G-SDI (5.94 Gbps, 2Kp30) and 12G-SDI (11.88 Gbps, 2Kp60).
Typically used in industrial cameras, MIPI-CSI-2 is a sensor interface that links image sensors to FPGAs. The PolarFire family now supports receive speeds up to 1.5 Gbps per lane and transmit speeds up to 1 Gbps per lane.
Customers can also implement a two-lane or eight-lane SLVS-EC Rx FPGA core to support high-resolution cameras at 2.3 Gbps per lane. The PolarFire family can support 1, 2.5, 5 and 10 Gbps speeds over an Ethernet PHY, enabling the initiative to meet the need for Universal Serial 10 GE Media Independent Interface (USXGMII) MAC IP with auto-negotiation.
Further IP such as the 6.25 Gbps CoaXPress v1.1 Host and Device IP can be used in high- performance machine vision, medical and in industrial inspection. Aligned with the industry’s roadmap for the standard, Microchip will support CoaXPress v2.0, which doubles the bandwidth to 12.5 Gbps.
The HDMI 2.0b IP core offered today supports resolutions up to 4K at 60 fps transmit and 1080p at 60 fps receive. The PolarFire FPGA Imaging IP bundle features the MIPI-CSI-2 and includes image processing IPs for edge detection, alpha blending and image enhancement for colour, brightness and contrast adjustments.
Last but not least, the Smart Embedded Vision partners ecosystem has been extended with the introduction of Kaya Instruments, which provides PolarFire FPGA IP Cores for CoaXPress v2.0 and 10 GigE vision, to Microchip’s partner ecosystem. The ecosystem also includes Alma Technology, Bitec and artificial intelligence partner ASIC Design Services, which provides a Core Deep Learning (CDL) framework that enables a power-efficient Convolutional Neural Network (CNN)-based imaging and video platform for embedded and edge computing applications.
With family members ranging from 100K to 500K Logic Elements (Les), the PolarFire FPGAs offer 30 to 50 percent lower total power over competing Static Random-Access Memory (SRAM)-based mid-range FPGAs, claims the manufacturer. Microchip has also unveiled a new MIPI-CSI2-based machine learning camera reference design for smart embedded system implementations. Based on the PolarFire FPGA imaging and video kit that uses inference algorithms from Microchip partner ASIC Design Services, the reference design is free for customers to evaluate.