NVIDIA IGX T5000 Industrial Edge AI Computing Module with Blackwell GPU Architecture
The NVIDIA IGX T5000 is a production-ready System-on-Module (SoM) designed for the most demanding industrial, medical, and mission-critical edge AI applications. Powered by the latest NVIDIA Blackwell GPU architecture, a 14-core Arm Neoverse V3AE processor, and a massive 128 GB of LPDDR5X memory, this US-made module delivers groundbreaking AI compute performance for real-time sensor processing, generative AI, and autonomous systems. Perfect for industrial automation, surgical robotics, autonomous mobile robots, and smart city infrastructure.
Unmatched AI Performance at the Edge
Built on NVIDIA Blackwell architecture, the IGX T5000 delivers exceptional AI inference performance with up to 2070 TFLOPS FP4 and 1035 TFLOPS FP8 compute capability. 5th-Generation Tensor Cores provide advanced AI cores with hardware support for FP4, FP8, FP16, bfloat16, TF32, and INT8 formats, enabling complex Large Language Models and deep learning workloads to run at the edge with ultra-low latency. The integrated Transformer Engine accelerates generative reasoning and multimodal sensor processing directly at the hardware level.
Key Features
-
Blackwell GPU Architecture: 2,560-core GPU with 96 fifth-generation Tensor Cores for breakthrough AI performance
-
High-Performance CPU: 14-core Arm Neoverse-V3AE 64-bit processor for real-time control and deterministic workloads
-
Memory & Bandwidth: 128 GB LPDDR5X with 273 GB/s memory bandwidth for massive AI models
-
Functional Safety Island: Dedicated safety processor on SoC designed to meet ISO 26262 and IEC 61508 standards
-
Flexible Power Envelope: Configurable 40 to 130 W power envelope for diverse deployment scenarios
-
Enterprise Software Support: 10-year product life cycle and enterprise software support from initial market availability
-
Compact Embedded Form Factor: System-on-Module design enables integration into custom carrier boards
-
Pin-Compatible Design: Form-factor compatible and pin-compatible with NVIDIA Jetson T5000
Industrial-Grade Reliability & Functional Safety
NVIDIA IGX is an industrial-grade edge AI platform that delivers high performance, advanced functional safety, and security, purpose-built for industrial and medical environments. The Functional Safety Island (FSI) inside the Thor SoC provides an independent safety processor that isolates safety-critical workloads, ensuring reliable operation in mission-critical environments. Multi-Instance GPU (MIG) support delivers robust workload isolation for running multiple AI models simultaneously.
Real-World Applications
-
Autonomous Mobile Robots & AGVs: Multi-camera and LiDAR data processing with hardware-accelerated depth mapping for warehouse automation
-
Industrial Quality Control: High-speed product defect detection and classification using multiple camera streams on production lines
-
Medical Imaging & Surgical Robotics: Real-time sensor processing and AI reasoning for advanced medical applications
-
Smart Cities & Security: Edge computing analytics systems capable of simultaneously processing 8K and multiple 4K video streams
-
Humanoid Robotics: Running generative AI models like NVIDIA Isaac GR00T for physical AI agents
Enterprise AI Software Stack
The IGX T5000 runs the complete NVIDIA AI Enterprise software suite, bringing cloud-native AI capabilities to the edge. Applications include NVIDIA Isaac for robotics, Metropolis for visual AI, Holoscan for sensor processing, and NVIDIA Halos safety AI agents. Supports NVIDIA NIM for deploying generative AI models including VLMs and LLMs at the edge.
Connectivity & Expandability
The module features flexible high-speed I/O for customization, enabling designers to create domain-specific industrial and robotic systems. Expanded high-speed I/O capabilities support diverse connectivity requirements including PCIe, USB, DisplayPort, Ethernet, and more through custom carrier board implementations.
What's in the Box
- NVIDIA IGX T5000 System-on-Module
- Product documentation
Full Specifications
Compute
| GPU Architecture |
NVIDIA Blackwell |
| GPU Cores |
2,560-core GPU with 96 fifth-generation Tensor Cores |
| AI Performance |
Up to 2,070 TFLOPS (FP4-Sparse), 1,035 TFLOPS (FP8) |
| CPU |
14-core Arm Neoverse-V3AE 64-bit processor |
| Memory |
128 GB 256-bit LPDDR5X |
| Memory Bandwidth |
273 GB/s |
Platform Features
| Form Factor |
System-on-Module (SoM) |
| Functional Safety |
Functional Safety Island on SoC, ISO 26262 & IEC 61508 ready |
| Power Consumption |
Configurable 40-130 W |
| Product Lifecycle |
10 years from initial availability |
| Compatibility |
Pin-compatible with NVIDIA Jetson T5000 |
| Manufacturing |
US Made |
Software Support
| Operating System |
NVIDIA IGX Software Stack (Linux-based) |
| AI Frameworks |
NVIDIA AI Enterprise, Isaac, Metropolis, Holoscan, NIM |
| Safety Software |
NVIDIA Halos OS for functional safety |
| AI Accelerators |
Transformer Engine, Multi-Instance GPU (MIG) |
Specifications are sourced from the manufacturer product page and may be updated.
