NVIDIA Jetson AGX Orin 64GB Developer Kit - 275 TOPS AI Edge Computing Platform
The NVIDIA Jetson AGX Orin 64GB Developer Kit delivers a giant leap forward for robotics and edge AI development. With up to 275 TOPS of AI performance and power configurable between 15W and 60W, this compact developer kit provides more than 8 times the performance of Jetson AGX Xavier in the same form factor. Perfect for prototyping advanced AI-powered robots, autonomous machines, and intelligent vision systems.
Why You'll Love the Jetson AGX Orin Developer Kit
-
Massive AI Performance: Up to 275 TOPS enables deployment of large, complex AI models for natural language understanding, 3D perception, and multi-sensor fusion
-
NVIDIA Ampere GPU Architecture: 2048 CUDA cores and 64 Tensor cores deliver professional-grade AI compute at the edge
-
Versatile Development Platform: Emulates all Jetson Orin modules including Orin NX and Orin Nano, letting you prototype multiple product configurations with one kit
-
Comprehensive Connectivity: PCIe Gen 4, 10GbE Ethernet, USB 3.2, M.2 NVMe, DisplayPort, and MIPI camera inputs provide extensive peripheral support
-
64GB Unified Memory: LPDDR5 memory with 204.8GB/s bandwidth feeds multiple concurrent AI pipelines
-
Complete Software Stack: Includes JetPack SDK with support for Isaac for robotics, DeepStream for vision AI, and Riva for conversational AI
Powerful AI Architecture for Edge Computing
At the heart of the developer kit is the Jetson AGX Orin 64GB module, featuring a 12-core Arm Cortex-A78AE CPU running at up to 2.2GHz, combined with the NVIDIA Ampere architecture GPU. The module includes dual NVIDIA Deep Learning Accelerators (NVDLA v2.0) and a Programmable Vision Accelerator (PVA v2.0) for hardware-accelerated AI inferencing. This powerful combination supports multiple concurrent AI application pipelines for computer vision, natural language processing, and sensor fusion applications.
Compact Form Factor, Maximum Connectivity
Despite its powerful specifications, the Jetson AGX Orin Developer Kit measures just 110mm × 110mm × 72mm, making it incredibly compact for desktop prototyping. The reference carrier board provides extensive I/O including:
- 2 × USB 3.2 Gen 2 Type-A ports
- 2 × USB 3.2 Gen 1 Type-A ports
- 2 × USB 3.2 Gen 2 Type-C ports with USB PD (one for power input)
- 1 × micro-USB 2.0 debug port
- 10 Gigabit Ethernet (10GbE) RJ45
- DisplayPort 1.4a output supporting up to 8K60
- PCIe Gen 4 x16 mechanical slot (x8 electrical)
- M.2 Key M slot (PCIe Gen 4 x4) for NVMe SSD storage
- M.2 Key E slot (PCIe Gen 4 x1) with pre-installed 802.11ac Wi-Fi and Bluetooth 5.0 module
- MicroSD card slot
- 120-pin MIPI CSI-2 camera connector supporting up to 6 cameras (16 via virtual channels)
- 40-pin GPIO expansion header with I2C, SPI, UART, CAN, I2S, PWM, and GPIO
- JTAG debug port
Ideal for Professional AI Development
The Jetson AGX Orin 64GB is purpose-built for demanding edge AI applications across industries including autonomous robotics, industrial automation, smart cities, healthcare imaging, retail analytics, and agricultural technology. The configurable power modes allow you to balance performance and thermal requirements for your specific application, while the comprehensive JetPack SDK provides optimised libraries for CUDA, TensorRT, cuDNN, and computer vision frameworks.
Complete Development Environment
NVIDIA provides extensive software support through the JetPack SDK, which includes Linux for Tegra (L4T) based on Ubuntu 22.04, CUDA toolkit, TensorRT for optimised inference, cuDNN for deep learning, VPI for computer vision, and multimedia APIs. Pre-trained models are available through the NGC catalog, and the NVIDIA TAO toolkit enables transfer learning and model fine-tuning. Omniverse Replicator supports synthetic data generation for training.
What's in the Box
- NVIDIA Jetson AGX Orin 64GB module with heatsink (pre-installed on carrier board)
- Reference carrier board (P3737)
- 802.11ac/abgn dual-band Wi-Fi and Bluetooth 5.0 module (pre-installed in M.2 Key E slot)
- 90W USB-C power adapter with international plugs
- Quick start guide
Full Specifications
Compute Performance
| AI Performance |
Up to 275 TOPS (INT8) |
| GPU |
2048-core NVIDIA Ampere architecture with 64 Tensor Cores, up to 1.3GHz |
| CPU |
12-core Arm Cortex-A78AE (ARMv8.2) 64-bit CPU, up to 2.2GHz, 3MB L2 + 6MB L3 cache |
| AI Accelerators |
2 × NVIDIA Deep Learning Accelerator (NVDLA) v2.0, 1 × Programmable Vision Accelerator (PVA) v2.0 |
| Memory |
64GB LPDDR5, 256-bit, 204.8GB/s bandwidth |
| Storage |
64GB eMMC 5.1 (onboard module) |
Connectivity & I/O
| Networking |
10 Gigabit Ethernet (10GbE), 802.11ac 2x2 Wi-Fi, Bluetooth 5.0 |
| USB Ports |
2 × USB 3.2 Gen 2 Type-A, 2 × USB 3.2 Gen 1 Type-A, 2 × USB 3.2 Gen 2 Type-C with USB PD, 1 × micro-USB 2.0 |
| Display Output |
1 × DisplayPort 1.4a (up to 8K60) |
| PCIe Expansion |
1 × PCIe Gen 4 x16 slot (x8 electrical), 1 × M.2 Key M (PCIe Gen 4 x4), 1 × M.2 Key E (PCIe Gen 4 x1) |
| Camera Input |
120-pin MIPI CSI-2 connector, 16-lane, D-PHY 2.1 (up to 40Gbps) or C-PHY 2.0 (up to 164Gbps) |
| Expansion Header |
40-pin GPIO with I2C, SPI, UART, CAN, I2S, PWM, GPIO |
| Other |
MicroSD card slot, JTAG debug connector |
Power & Physical
| Power Input |
USB Type-C (USB PD) or DC barrel jack (9-20V) |
| Power Consumption |
Configurable 15W to 60W |
| Dimensions |
110mm × 110mm × 72mm (4.3" × 4.3" × 2.8") |
| Operating Temperature |
0°C to 50°C |
Software
| Operating System |
Linux for Tegra (L4T) based on Ubuntu 22.04 LTS |
| SDK |
JetPack 5.x and later |
| AI Frameworks |
TensorFlow, PyTorch, MXNet, ONNX Runtime via NGC containers |
| AI Tools |
TensorRT, cuDNN, CUDA Toolkit, DeepStream SDK, Isaac SDK, Riva SDK |
| Computer Vision |
VPI (Vision Programming Interface), OpenCV |
| Multimedia |
Hardware-accelerated video encode/decode (H.265, H.264, VP9) |
Specifications are sourced from the official NVIDIA developer documentation and may be updated. Visit developer.nvidia.com for the latest information.
