Taipei, Taiwan — May 2026 — Lambortech today announced the launch of the ECM400-Thor, a next-generation Edge AI computing platform designed for robotics, autonomous systems, and advanced physical AI applications. Powered by NVIDIA® Jetson Thor™ T5000/T4000 modules, the compact platform delivers up to 2070 FP4 TFLOPS AI performance in a highly integrated embedded design optimized for real-time perception, sensor fusion, and AI-driven decision-making.

Built for emerging robotics workloads, the ECM400-Thor combines NVIDIA Blackwell GPU architecture with rich high-speed connectivity, flexible expansion, and multi-sensor integration capabilities. The system supports advanced robotics applications including humanoid robots, AMR/AGV platforms, autonomous machines, intelligent inspection systems, and edge-based generative AI inference.

The ECM400-Thor features up to 128GB LPDDR5X memory, 16× MIPI CSI-2 lanes for high-bandwidth camera integration, 5GbE networking, multiple M.2 expansion interfaces, CAN Bus support, and customizable high-speed PCIe Gen5 expansion through dedicated connectors. Integrated IMU sensors and optional RTK/GNSS support further enhance deployment flexibility for robotics navigation and autonomous mobility systems.

"With robotics and physical AI rapidly evolving, next-generation autonomous systems require significantly higher AI performance, faster sensor processing, and scalable edge computing architectures," said Lambortech. "The ECM400-Thor is designed to help developers accelerate advanced robotics innovation with powerful AI computing, compact integration, and flexible peripheral expansion."

Powered by NVIDIA Jetson Thor™, the platform supports real-time execution of large AI models, transformer-based AI workloads, multi-modal perception, and advanced robotics frameworks. NVIDIA positions Jetson Thor as its most powerful embedded AI platform for physical AI and humanoid robotics, offering over 7.5× higher AI compute performance compared to Jetson AGX Orin.

Key Features of ECM400-Thor

• Powered by NVIDIA® Jetson Thor™ T5000 / T4000
• Up to 2070 FP4 TFLOPS AI performance
• Up to 128GB LPDDR5X memory
• 16× MIPI CSI-2 lanes for multi-camera integration
• High-speed PCIe Gen5 expansion and 5GbE LAN
• Designed for robotics and physical AI applications

For more information about ECM400-Thor, visit:
ECM400-Thor Product Page