Gidel, an Israel-based tech company specialised in high-speed image acquisition and processing, has announced that the new NVIDIA Jetson Orin NX 16GB system-on-module (SoM) will be added to Gidel’s FantoVision range of hi-bandwidth compact computers for vision on the edge.
The new FantoVision with the NVIDIA Jetson Orin edge AI platform will be presented at the Embedded World show in Nuremberg, Germany, on March 14-16 2023.
With up to 100 sparse TOPS, the Jetson Orin NX 16GB module increases the AI processing power of FantoVision computers by a factor of 5.
Paired with Gidel’s high-speed camera interfaces and FPGA-based frame grabbing and processing technology, the new processor opens unprecedented possibilities for real-time, low-latency image processing on the edge for industrial machine vision and other applications. It includes 10GigE, CoaXPress 12 and Camera Link interfaces for image acquisition, processing and compression at over 1 gigapixel per second.
Gidel’s FantoVision edge computers have been specially designed to make NVIDIA Jetson modules easily accessible to high-end imaging and vision system developers. They combine NVIDIA Jetson processors with high-bandwidth camera interfaces and Gidel’s acclaimed FPGA-based frame grabbing and acceleration technology. The result is a ready-to-use, compact, low-power industrial computer that fulfills the highest image acquisition standards.
With Gidel’s high-bandwidth interfaces, including 10GigE, CoaXPress 12 and Camera Link, FantoVision computers can capture and process images at up to 40 gigabits per second for high-resolution, high-speed, or multi-camera applications.
Unique platform for AI machine vision
The NVIDIA Jetson Orin NX 16GB module delivers five times more performance than the previous-generation Jetson Xavier NX platform with 100 TOPS vs. 21 TOPS, respectively. It also features a more powerful, new-generation CPU and GPU. With this level of performance, it can process the high image data throughput delivered by the Gidel frame grabber in real time at more than 1 gigapixel per second using the NVIDIA JetPack SDK.
This makes FantoVision a unique solution for machine vision system designers who want to run a deep learning model on the edge, without compromising on real-time processing capabilities.
“The combination of Gidel’s FPGA and image processing know-how and NVIDIA accelerated computing breaks another barrier between AI and machine vision,” said Reuven Weintraub, Founder and CTO of Gidel. “With our NVIDIA Jetson Orin-based FantoVision edge computers, vision system engineers have a turn-key solution to implement AI in machine vision systems.”
