NVIDIA Jetson Thor Module
Original price was: ₹1,299,999.00.₹749,999.00Current price is: ₹749,999.00.
The NVIDIA Jetson Thor Module is a next-generation embedded AI platform powered by the Blackwell GPU architecture and a Grace-based ARM CPU, delivering up to 2,070 FP4 TFLOPS of AI performance. With support for JetPack SDK, CUDA, TensorRT, DeepStream, and Isaac ROS, it is designed for humanoid robots, autonomous machines, industrial automation, computer vision, and advanced edge AI applications.
NVIDIA Jetson Thor Module
Features
- Powered by the NVIDIA Blackwell GPU architecture.
- Delivers up to 2,070 FP4 TFLOPS of AI performance (configuration dependent).
- High-performance Grace-based ARM CPU for robotics and AI applications.
- Up to 128GB LPDDR5X unified memory with high memory bandwidth.
- Supports large vision-language models (VLMs), transformers, and generative AI.
- Native support for CUDA, TensorRT, cuDNN, DeepStream, Isaac ROS, and JetPack SDK.
- Extensive I/O including PCIe Gen5, USB4, MIPI CSI, CAN, GPIO, and more.
- Hardware acceleration for AV1, H.265, and H.264 video processing.
- Designed for humanoid robots, autonomous mobile robots (AMRs), industrial automation, smart cities, and edge AI.
- Compatible with NVIDIA’s robotics and AI software ecosystem.
Specifications
| Product Name | NVIDIA Jetson Thor Module |
| Brand | NVIDIA |
| Series | NVIDIA Jetson™ |
| Model | Jetson Thor Module |
| Product Type | Edge AI Compute Module |
| AI Architecture | NVIDIA Blackwell GPU + Grace-Class CPU Architecture |
| GPU Architecture | NVIDIA Blackwell |
| CPU | NVIDIA Grace-based ARM CPU |
| AI Performance | Up to 2,070 FP4 TFLOPS (Model Dependent) |
| CUDA Cores | Blackwell CUDA Cores with 5th Gen Tensor Cores |
| Tensor Cores | 5th Generation Tensor Cores |
| Memory | Up to 128GB LPDDR5X Unified Memory |
| Memory Bandwidth | Up to 273 GB/s |
| Storage Support | NVMe SSD (via Carrier Board) |
| Camera Support | Multiple MIPI CSI-2 Cameras |
| Display Support | DisplayPort / HDMI (Carrier Board Dependent) |
| Networking | 10Gb Ethernet, Wi-Fi, 5G (Carrier Board Dependent) |
| Expansion Interfaces | PCIe Gen5, USB4, CAN, UART, SPI, I²C, GPIO |
| Video Encode/Decode | Hardware AV1, H.265 (HEVC), H.264 |
| Operating System | NVIDIA JetPack SDK (Ubuntu Linux) |
| AI Frameworks | CUDA, TensorRT, cuDNN, DeepStream, Isaac ROS, TAO Toolkit |
| Power Modes | Configurable (Application Dependent) |
| Target Applications | Robotics, Humanoid Robots, Autonomous Machines, Industrial AI, Edge A |





