Neousys Technologies has launched the NRU-160-FT, a compact computing system designed for installation in sealed enclosures for various applications, including construction machinery and autonomous logistics vehicles, where traditional air-cooled computers may not be effective. The system leverages the enclosure walls for heat dissipation, enabling operation without active cooling, a key advantage for rugged environments.
Built on Nvidia’s Jetson Orin platform, the NRU-160-FT is offered with either the Orin NX or Orin Nano module. It boasts an impressive capability, delivering up to 100 INT8 tera-operations per second (TOPS) of artificial intelligence performance. The device can simultaneously process as many as 18 Full HD video streams while consuming approximately 25 watts of power. Utilizing CUDA and Tensor cores, it facilitates local neural network execution, making it suitable for demanding AI tasks.
Measuring 199 × 136 × 49 mm, the NRU-160-FT is aimed at developers and integrators in sectors such as robotics, automation, automotive, and transport technology. Its fanless design employs conduction cooling, transferring heat directly to a mounting surface through a flat passive cooling profile. Neousys states that the system can operate in ambient temperatures ranging from –25 °C to +70 °C without any clock speed reduction when configured for 25 W operation, further enhancing its viability for outdoor and industrial applications.
The NRU-160-FT family includes several models tailored for specific camera and sensor requirements. The NRU-161V-FT can support up to six GMSL2 cameras, making it particularly useful in automotive and driver-assistance scenarios. Pre-installed drivers for various Sony image sensors, including the IMX390, IMX490, and ISX031, enable applications in safety systems and robotic navigation.
Another variant, the NRU-162S-FT, serves the needs of IP and industrial camera applications, providing five Gigabit Ethernet interfaces, four of which support Power over Ethernet (PoE). All connections utilize M12 connectors to enhance reliability in mobile and harsh environments, ensuring sustained performance even under challenging conditions.
Power input for the NRU-160-FT is flexible, supporting a range from 8 to 35 V, making it suitable for deployment in vehicles and mobile machinery. The system includes power-on control features that allow it to automatically start with a vehicle and safely shut down when the ignition is turned off, a necessary capability for many industrial and transport applications. The absence of fans or filters also reduces maintenance demands, contributing to lower overall energy consumption and extended battery life.
Additional interfaces include CAN-FD, RS-232, mini-PCIe, and an M.2 B-Key slot for mobile communication modules, including support for cellular connectivity up to 5G. This extensive connectivity enhances the system’s adaptability for various applications, including real-time data processing and remote monitoring.
Acceed, the distributor of the NRU-160-FT, emphasizes that the system is designed to facilitate the transition from laboratory development to long-term deployment in real-world environments, particularly in scenarios where reliable edge AI processing is essential. The company is providing the system in multiple configurations along with support for project planning, installation, and ensuring long-term availability.
As industries increasingly embrace automation and AI technologies, the NRU-160-FT represents a significant step toward enabling more efficient and intelligent systems in challenging environments. Its combination of robust performance, compact design, and low maintenance requirements positions it well for the evolving landscape of edge AI applications.
See also
Sardden Launches Optimized Token to Democratize AI Power in Decentralized Networks
Caason Group Advances AI-Driven Manufacturing Strategy Under Craig Astill’s Leadership
Google Highlights Taiwan’s Central Role in AI Transition, Unveils New ‘Personal Intelligence’ Feature
Tesla Reveals AI6 and AI6.5 Chips: Samsung and TSMC to Manufacture 2nm Designs in 2027-2029
AI Adoption Accelerates: 54% of SMEs Report Productivity Gains, Outpacing Competitors
