New Ethernovia HSB platform accelerates real-time sensor data streaming and edge AI workloads on NVIDIA Holoscan systems

Ethernovia, Inc., a leader in high-performance Ethernet networking and packet-processing silicon for physical AI systems, today announced the launch of its new HSB-compatible networking platform designed to integrate seamlessly with NVIDIA Holoscan Sensor Bridge (HSB)-enabled edge AI platforms. The Ethernovia HSB platform brings deterministically low-latency, high-bandwidth sensor data transport over Ethernet to NVIDIA Holoscan deployments, while extending data pipelines into real-time simulations and digital twins built on NVIDIA Omniverse libraries.

Ethernovia will be showcasing a live demo of its Ethernet Sensor Bridge (HSB) platform at the NVIDIA GTC 2026 Conference this week taking place in the San Jose Convention Centre.

Building on Ethernovia’s expertise in Ethernet-based packet processors — developed to serve as the real-time data backbone for next-generation physical AI systems — the Ethernovia HSB platform provides a scalable, software-defined Ethernet backbone that transports high-bandwidth, multi-sensor data directly into GPU memory on NVIDIA Holoscan platforms — enabling real-time AI perception, control, and inference at the edge. By leveraging Ethernet as a unified networking fabric, the board simplifies system architecture while supporting synchronized sensor streams critical for robotics, autonomous systems, industrial automation and medical imaging.

“Physical AI systems demand a networking foundation that is deterministic, scalable, and software-defined,” said Ramin Shirani, CEO and co-founder of Ethernovia. “Our HSB-enabled platform delivers real-time Ethernet sensor connectivity optimized for NVIDIA Holoscan while extending seamlessly into digital twins built on Omniverse libraries. This creates a unified pipeline from live sensor capture to simulation and AI refinement — dramatically accelerating the development of autonomous and intelligent machines.”

Beyond real-time deployment, Ethernovia’s Ethernet architecture enables seamless streaming of live and recorded sensor data into applications built on NVIDIA Omniverse libraries, enabling digital twin modeling, validation and closed-loop simulation. This tight coupling between sensor aggregation (HSB), AI compute (Orin / Thor) and simulation allows developers to:

• Validate perception stacks against real sensor traffic
• Replay deterministic Ethernet data streams into simulation environments
• Accelerate model training and system verification
• Reduce integration cycles between lab, simulation, and field deployment

Ethernovia’s new HSB platform is now available for evaluation, details can be found at www.ethernovia.com/products/physical-ai-platforms/ethernet-sensor-bridge-for-nvidia-holoscan/

Additional Product Information

For additional product information and to arrange a meeting / demo at GTC, please contact: sales@ethernovia.com.

About Ethernovia

Ethernovia is developing the future of Ethernet-based networks to realize the full potential of software-defined and autonomous vehicles, robotics and other intelligent machines. The company’s breakthrough data transport and acceleration technology is ushering in a new era of connectivity and capabilities in the vehicle and at the edge, including highly reliable autonomy, over-the-air servicing and AI-backed applications. To learn more, please visit www.ethernovia.com and follow Ethernovia on LinkedIn.

View source version on businesswire.com: https://www.businesswire.com/news/home/20260316729719/en/