← Back
NVIDIA
NVIDIA launches IGX Thor edge AI platform for industrial and medical robotics with Blackwell compute
· releaseplatformedge-aifeature · developer.nvidia.com ↗

Overview

NVIDIA IGX Thor is a new enterprise-ready platform purpose-built for deploying high-performance AI at the industrial and medical edge. The platform combines server-class compute capabilities with industrial-grade hardware, advanced functional safety, extended lifecycle support, and an enterprise software stack designed for mission-critical deployments where uptime, reliability, and standards compliance are essential.

Platform Configuration

The IGX Thor family consists of four purpose-built configurations:

  • IGX T5000 System-on-Module (SoM): Compact embedded form factor delivering high-performance, safety-capable compute for integration into custom carrier boards. Optimized for building domain-specific industrial and robotic systems with accelerated time-to-market.
  • IGX T7000 Board Kit: High-performance variant on MicroATX form factor scaling for demanding workloads. Includes rich I/O, optional discrete GPU support (RTX PRO 6000 or 5000 Blackwell), and enterprise networking capabilities.
  • IGX Thor Developer Kit: Full-featured development platform with advanced sensing, robotics support, and real-time AI pipeline capabilities for building and validating next-generation industrial AI applications.
  • IGX Thor Developer Kit Mini: Space and power-constrained variant ideal for mobile robots and compact industrial systems that require robust AI performance in smaller footprints.

Key Technical Specifications

All variants feature Blackwell GPU architecture with 2,560-core integrated GPUs and 14-core Arm Neoverse-V3AE CPUs. Performance ranges from 2,070 FP4 TFLOPS (T5000 and Mini Kit) to 5,581 FP4 TFLOPS (T7000 and full Developer Kit). All configurations include 128 GB LPDDR5x memory and multi-instance GPU support with 10 TPCs. The T7000 and Developer Kit support discrete GPU additions for enhanced performance.

Safety and Compliance

A critical differentiator is hardware-based functional safety with dedicated Functional Safety Islands (FSI) integrated into the SoC, plus dedicated Safety MCUs in higher-end variants. These features enable compliance with ISO 26262 and IEC 61508 standards essential for regulated industrial and medical environments.

Networking and I/O

Enterprise-grade connectivity includes dual 200 GbE networking with RDMA support on T7000 and Developer Kit models for low-latency sensor data handling. Lower-end variants feature 4x 25 Gbps MGBE ports. All configurations include USB-C, NVMe SSD storage, and optional Wi-Fi 6E connectivity.

Use Cases

The platform targets factory automation, autonomous mobile platforms, surgical systems, and other safety-critical edge AI applications requiring deterministic behavior, high availability, and verifiable functional safety.