Advantech Introduces Next-Gen Edge AI System Powered by NVIDIA IGX Thor for Real-Time, Safe Intelligence

Advantech, Taipei, Taiwan, 17.03.2026 - Advantech, a global leader in industrial edge computing, today announced the expansion of its edge AI portfolio powered by NVIDIA IGX Thor, targeting next-generation industrial and medical edge applications, including advanced medical imaging, and AI-assisted diagnostics. The new portfolio is built on NVIDIA IGX Thor, a powerful, industrial-grade platform designed to bring real-time physical AI directly to the edge by combining high-speed sensor processing, robust reliability, and functional safety – enabling intelligent systems to operate with deterministic performance in mission-critical environments. 

As AI moves beyond perception into real-world decision making, NVIDIA IGX Thor provides the foundation for safety-critical, low-latency AI inference at the edge, bringing physical AI from concept to deployment.

NVIDIA IGX Thor: The Foundation for Physical AI at the Edge

NVIDIA IGX Thor is designed to meet the stringent demands of physical AI applications, where deterministic performance, functional safety, and system reliability are critical requirements for deployment. Built for industrial and medical-grade environments, the platform combines high-performance AI compute, real-time processing, and safety capabilities in a unified architecture, forming a solid foundation for safety-critical edge intelligence.

Building on NVIDIA IGX Thor, Advantech delivers an industrial-grade edge AI portfolio designed for long-term deployment in industrial and medical environments. With a 10-year lifecycle and support for the NVIDIA AI Enterprise - IGX software stack, Advantech enables customers to accelerate physical AI applications from cloud to edge using NVIDIA NIM microservices, NVIDIA Isaac foundation models and runtime libraries for robotics, NVIDIA Metropolis for visual AI, and NVIDIA Holoscan for real-time sensor processing.

By integrating capabilities from the NVIDIA Halos AI Systems Inspection Lab, Advantech embeds functional safety into its platforms based on NVIDIA IGX Thor, supporting safe human – machine collaboration across robotics, medical, and industrial AI systems operating in real-world conditions.

Outstanding Edge AI Solutions for Deployment at Scale

Advantech structures its offerings built on NVIDIA IGX Thor as a portfolio of industrial and medical edge platforms designed to support different deployment scenarios. Within this portfolio, Advantech offers multiple product categories, including board-level platforms for development and customization, as well as system-level platforms optimized for industrial deployment. This approach allows customers to apply a consistent AI architecture across applications while selecting platforms that align with their integration, performance, and safety needs.

MIC-735 Industrial AI Inference System

Powered by NVIDIA IGX T5000 module, MIC-735 is a production-ready edge AI inference system delivering up to 2,070 FP4 TFLOPS with functional safety MCU integration for deterministic AI in safety-critical environments. By integrating NVIDIA Holoscan platform including Advantech’s Holoscan Sensor Bridge, MIC-735 enables high-bandwidth, synchronized communication between multiple sensors and AI processors, supporting coordinated perception and control for robotics and medical systems. NVIDIA Halos Outside-In Safety Agents running on IGX Thor analyze footage from cameras installed in the surrounding environment to extend a robot’s field of vision beyond its onboard sensors.

In addition to functional safety hardware, the MIC-735 supports NVIDIA Halos AI Systems Inspection Lab, incorporating both onboard and external sensor data to enhance safe human–machine interaction.

To meet diverse integration needs, Advantech also offers the MIB-735, a board-level platform based on the same NVIDIA IGX Thor architecture. With dedicated thermal design, it enables customized mechanical integration for space-constrained applications.

AIR-427A Medical-Grade Edge AI HPC

Built on NVIDIA IGX Thor, the AIR-427A is a medical-grade HPC platform engineered for safety-critical healthcare environments, delivering deterministic and high-reliability performance for clinical AI workloads. Featuring dual NVIDIA RTX PRO 6000 Blackwell Max-Q GPUs and up to 320GB of total VRAM, it delivers exceptional compute power for real-time medical imaging, LLM-assisted diagnostics, and multi-modal data processing. Advantech strengthens the platform with its Edge AI SDK, system validation services, and security management to streamline deployment and support long-term lifecycle operations.

By integrating NVIDIA’s safety-focused architecture with Advantech’s system and software expertise, the AIR-427A enables secure, low-latency AI inference at the edge for scalable and compliant medical deployments.

USM-500 IGX Thor Medical-Grade AI Computer

Designed for advanced imaging, surgical, and diagnostic systems, USM-500 delivers high-performance medical AI computing for mission-critical environments. Built on NVIDIA IGX Thor and accelerated by NVIDIA RTX PRO 6000 Blackwell Max-Q, it provides up to 5,581 FP4 TFLOPs (Sparse), with integrated NVIDIA ConnectX-7 dual 100GbE networking for real-time, high-bandwidth clinical workloads.

Aligned with IEC 60601 standards and validated for medical environments, USM-500 ensures deterministic performance and long lifecycle support. It is ideal for applications including CT analysis, surgical visualization, AI-assisted diagnostics, and high-throughput medical imaging.

Advancing the IGX Thor Ecosystem

Beyond individual systems, Advantech is committed to expanding the NVIDIA IGX Thor ecosystem through close collaboration with camera vendors, sensor partners, and AI software providers. This ecosystem-driven approach ensures that platforms built on NVIDIA IGX Thor are not only powerful, but also interoperable, validated, and ready for real-world applications.

Through this collaboration, Advantech and NVIDIA are together accelerating the adoption of physical AI, bringing safe, real-time intelligence to the edge.