NVIDIA Jetson AGX Orin 64GB AI Computing Module
The NVIDIA Jetson AGX Orin 64GB module delivers groundbreaking AI performance for next-generation robotics, autonomous machines, and edge AI applications. With up to 275 TOPS of AI performance and power configurable between 15W and 60W, this system-on-module provides more than 8X the performance of Jetson AGX Xavier in the same compact form-factor. Powered by the NVIDIA Ampere architecture, it features a 2048-core GPU with 64 Tensor cores, a 12-core Arm Cortex-A78AE 64-bit CPU, and 64GB of LPDDR5 memory with 204.8GB/s bandwidth, enabling multiple concurrent AI pipelines and complex model deployment.
Unmatched AI Performance for Robotics and Edge Computing
Built for developers and engineers working on advanced autonomous systems, the Jetson AGX Orin 64GB module supports demanding tasks including natural language understanding, 3D perception, multi-sensor fusion, and real-time vision AI. With 2x NVDLA v2.0 deep learning accelerators and PVA v2.0 vision accelerators, it delivers the compute power needed for industrial robotics, smart cities, autonomous vehicles, drones, healthcare devices, and more.
Key Features
-
275 TOPS AI performance: Up to 170 Sparse TOPS of INT8 Tensor compute and 5.3 FP32 TFLOPS of CUDA compute
-
2048 CUDA cores & 64 Tensor cores: NVIDIA Ampere architecture GPU with next-generation AI acceleration
-
12-core Arm Cortex-A78AE CPU: 64-bit processing with 3MB L2 and 6MB L3 cache, up to 2.2GHz
-
64GB LPDDR5 memory: 256-bit interface with 204.8GB/s bandwidth for high-throughput AI workloads
-
64GB eMMC 5.1 storage: Fast onboard storage for operating systems and applications
-
Configurable power modes: 15W to 60W for flexibility across deployment scenarios
-
Compact system-on-module design: 100mm × 87mm form-factor, pin-compatible with Jetson AGX Xavier
-
Advanced accelerators: 2x NVDLA v2.0 deep learning engines and PVA v2.0 vision processor
Comprehensive Software Ecosystem
The Jetson AGX Orin runs the complete NVIDIA AI software stack, including JetPack SDK, Isaac for robotics, DeepStream for vision AI, and Riva for conversational AI. Developers can leverage NVIDIA Omniverse Replicator for synthetic data generation, TAO toolkit for fine-tuning pretrained models from the NGC catalog, and full CUDA, TensorRT, and cuDNN support for accelerated inferencing. An extensive ecosystem of partners provides additional AI software, cameras, sensors, carrier boards, and custom development services.
Ideal Applications
- Autonomous mobile robots (AMRs) and collaborative robots
- Industrial automation and machine vision systems
- Smart city infrastructure and edge analytics
- Autonomous vehicles and drones
- Healthcare imaging and diagnostic devices
- Retail automation and service robots
- Agriculture and precision farming equipment
- Multi-camera surveillance and video analytics
Designed for Production Deployment
The Jetson AGX Orin 64GB module is a production-ready system-on-module engineered for integration into commercial products. Its compact form-factor, flexible power modes, and compatibility with carrier boards make it ideal for developing and deploying advanced AI solutions at scale. With support for high-speed PCIe, USB 3.2, MGBE Ethernet, MIPI CSI camera interfaces, and more, it provides the connectivity needed for real-world edge AI applications.
What's in the Box
- NVIDIA Jetson AGX Orin 64GB Module
Full Specifications
AI & GPU
| AI Performance |
275 TOPS (INT8) |
| GPU Architecture |
NVIDIA Ampere |
| CUDA Cores |
2048 |
| Tensor Cores |
64 (3rd generation) |
| GPU Max Frequency |
1.3 GHz |
| FP32 Performance |
5.3 TFLOPS |
CPU
| CPU |
12-core Arm Cortex-A78AE v8.2 64-bit |
| CPU Max Frequency |
2.2 GHz |
| CPU Cache |
3MB L2 + 6MB L3 |
Memory & Storage
| Memory |
64GB 256-bit LPDDR5 |
| Memory Bandwidth |
204.8 GB/s |
| Storage |
64GB eMMC 5.1 |
Accelerators
| DL Accelerator |
2x NVDLA v2.0 (1.6 GHz max frequency) |
| Vision Accelerator |
PVA v2.0 |
Power
| Module Power |
15W - 60W (configurable) |
| Voltage Input |
5V, 7V-20V |
Physical
| Form Factor |
100mm × 87mm |
| Module Type |
System-on-Module (SOM) |
Specifications are sourced from the manufacturer product page and may be updated.
