Neousys Technology, an industry-leading provider of rugged embedded systems, has announced the Nuvo-10109GC, a new high-performance edge AI platform built to support NVIDIA’s 600W Blackwell GPU. The system delivers up to 4000 AI TOPS for FP4 LLM inference and generative AI, or 125 TFLOPS for computer vision and autonomous workloads. It is engineered for deployment in critical environments, offering robust power delivery, flexible expansion for diverse I/O and sensing requirements, and is MIL-STD-810H compliant, with GPUs installed for mechanical reliability.

Neousys Nuvo-10109GC integrates a new power architecture capable of sustaining full output for a 600W GPU even at 9V low-voltage input, ensuring stable operation in vehicles and field installations where fluctuating power conditions are common. Its wide 9V–32V DC input range enables seamless compatibility with 12V and 24V automotive power systems, making it easy to deploy for in-vehicle applications. Engineered for rugged environments, it incorporates a reinforced chassis, an upgraded GPU retention mechanism, and vibration-damping hardware to ensure structural reliability under shock, vibration, and continuous on-road operation. The GPU retention brackets are specifically designed to secure today’s GPUs, often weighing 2kg or more, while still complying with MIL-STD-810H standards.

With a rich set of high-bandwidth I/O expansion interfaces, it can handle the massive data streams required by modern AI models and multi-sensor fusion. It features 10GbE and 1GbE Ethernet ports for communication, a front-accessible 2.5″ SATA drive bay, and an optional M.2 2280 Gen4x4 NVMe slot for real-time data recording. Additionally, it provides three extra Gen3x4 PCIe slots beyond the GPU slot, allowing system integrators to add GMSL2 camera capture cards, automotive Ethernet interfaces, high-resolution frame grabbers, or other domain-specific modules.

Combined with its electrical resilience, mechanical durability, and extensive expansion headroom, Nuvo-10109GC marks a major milestone for edge AI computing—offering performance, stability, and scalability required for next-generation AI deployments across both stationary edge nodes and demanding mobile environments.