Now a power upgrade to this rugged computer means that developers can benefit from additional flexibility when developing generative AI models.
The RSA4NA, which is based on NVIDIA Jetson Orin Nano system-on-modules (SoMs), is a platform that powers AI applications and as a specialist in rugged computing, Syslogic has ensured that the platform can be used in challenging environmental conditions.
According to Syslogic, the RSA4NA is now able to unlock new opportunities for generative AI models. Following a software update, the module’s maximum power consumption has increased from 15 watts to 25 watts enabling the computer’s GPU to operate at up to 1020 MHz, delivering a peak AI performance of 67 TOPs.
With this performance boost, sophisticated generative AI models such as large language models (LLMs), vision language models (VLMs) and vision transformers can now be run more efficiently than before on these compact devices.
The RSA4NA features an IP67 and IP69 rated housing, providing protection against moisture, water, and dust. Syslogic pairs the Jetson Orin Nano modules with a custom-designed carrier board to achieve exceptional shock and vibration resistance. Additionally, the edge computer is built to operate reliably in temperatures ranging from –40 to +70 Celsius.
The Rugged Computer RSA4NA not only provides exceptional performance and durability but also with its RTK (Real-Time Kinematic) technology it can offer real-time, centimetre-accurate positioning of vehicles or machines, making it an essential feature for autonomous mobile robots and vehicles.
Syslogic has leveraged u-blox technology so the RSA4NA can be equipped with two integrated u-blox receivers, enabling RTK functionality to be implemented quickly and without additional hardware. As an added feature, Syslogic has incorporated a heading function, allowing precise determination not only of a vehicle or machine’s position but also its angle of rotation.
Depending on the interface configuration, the RSA4NA supports up to four GMSL2 cameras, with image data processed virtually in real time. The cameras are powered directly via POC (Power over Coaxial), making the edge computer suitable for applications such as safety monitoring around machines or 360-degree bird's-eye view visualisation.
Featuring NVIDIA Jetson modules, RTK functionality, GMSL2 interfaces, and IP67/IP69 protection, this compact embedded computer provides a hardware platform for advanced computer vision, perception, and sensor fusion applications.
