Latest AI-enabling frame grabber cards are available now

29-06-2023 | Impulse Embedded | Industrial

Impulse Embedded has introduced the latest in AI-enabled frame grabbers from Neousys Technology. Two cards, the PCIe-NX156U3 and the PCIe-NX154PoE, are powered by the NVIDIA Jetson Orin NX SoM and with the option of either the six-port USB or a four-port PoE+ camera interface, they are a turnkey industrial frame grabber solution for AI-based vision applications.

Excellent for developers of machine vision and automated inspection systems, these PCI Express cards make it straightforward to add the power of Nvidia Jetson to 19" rackmount and x86 Industrial Computer platforms, shifting the AI processing load from the host computer and onto the card. These cards can also make a terrific addition to existing systems by upgrading them to the latest image capture, pre-processing, and inference technology, particularly considering they are created to withstand harsh environments by supporting a wide operating temperature of - 25C to +60C.

Thanks to NVIDIA's Jetpack software suite, the application software can be developed on any Jetson-powered device and then seamlessly deployed onto these capture cards, accelerating time to market for machine vision integrators already fluent with this software suite. Moreover, these cards can function in standalone mode, so developers can produce their applications before installing them into a host system.

The PCIe-NX156U3 has a USB interface offering six USB3.2 ports, each with 10Gbps bandwidth and up to 1500mA current for camera connectivity, while the PCIe-NX154PoE has four high-speed 2.5Gbps PoE+ ports, supporting a total of 50W for IP cameras, with the cards having software-programmable per-port power control.

The standard, single-slot, half-length cards use a PCIe x4 slot with an x1 signal, which is compatible with Windows and Linux systems. These AI-enabling frame grabber cards offload deep learning vision computing from the host system, which is ideal for image processing applications such as object sorting, defect detection, packaging inspection, assembly verification, and robotic guidance.

By Seb Springall