NVIDIA Jetson AGX Orin 64GB 개발자 키트
(NVIDIA Jetson AGX Orin 64GB Developer Kit)
개요
- 본 제품은 NVIDIA Jetson AGX Orin 64GB 개발자 키트입니다.
- 컴팩트한 크기를 가지고 있으며 최대 275 TOPS의 AI 성능을 가지고 있어 인공지능 탑재 로봇이나 자율 머신을 만드는데 적합합니다.
- multiple concurrent AI application pipelines, NVIDIA Ampere GPU architecture, next-generation deep learning , vision accelerator 를 지원합니다.
- NVIDIA AI 소프트웨어 스택 및 IsaacTM for robotics, DeepStream for vision AI, Riva for conversational AI 와 같은 어플리케이션 프레임워크를 사용할 수 있습니다.
특징
-
The NVIDIA Jetson AGX Orin 64GB Developer Kit makes it easy to get started with Jetson Orin. Compact size, lots of connectors, and up to 275 TOPS of Al performance make this developer kit perfect for prototyping advanced Al-powered robots and other autonomous machines.
The developer kit includes a Jetson AGX Orin 64GB module, and can emulate all the Jetson Orin modules. It supports multiple concurrent Al application pipelines with the NVIDIA Ampere GPU architecture, next-generation deep learning and vision accelerators, high-speed IO, and fast memory bandwidth. Now you can develop solutions using your largest and most complex Al models to solve problems such as natural language understanding, 3D perception, and multi-sensor fusion.
NVIDIA Jetson AGX Orin 64GB Developer Kit runs the entire Jetson software stack,as well as use-case specific application frameworks, including lsaac for robotics,DeepStream for vision A, and Riva for conversational Al. You can save significant time with NVIDIA Omniverse Replicator for synthetic data generation (SDG), and by using NVIDIA TAO toolkit to fine-tune pre-trained Al models from the NGC" catalog.
Jetson ecosystem partners offer additional Al and system software, developer tools, and custom software development. Ecosystem partners also offer cameras and other sensors for Jetson, as well as carrier boards and design services for your product.
Get the computing capability of more than eight Jetson AGX Xavier systems in a developer kit that integrates the latest NVIDIA GPU technology with the world's most advanced deep learning software stack. Now, you'll have the flexibility to create the next generation of Al solutions.
-
JetPack SDK
The JetPack SDK is compatible with all Jetson modules and developer kits. It is a comprehensive solution for building AI applications and provides a robust software foundation for Jetson AGX Orin. It includes the latest Linux driver package (L4T) for the Linux operating system, as well as the CUDA-X acceleration libraries and APIs for deep learning, computer vision, accelerated computing, and multimedia.
- TensorRT and cuDNN are used for high-performance deep learning applications.
- CUDA-X is a CUDAGPU acceleration application that spans multiple domains.
- NVIDIA Container Runtime is used for containerized GPU-accelerated applications.
- The multimedia API package is used for camera and sensor driver development.
- VisionWorks, OpenCV and VPI are used for visual computing.
- Examples of various demo applications.
- GPU: NVIDIA Ampere architecture with 2048 NVIDIA CUDA cores and 64 tensor cores
- CPU: 12-core Arm Cortex-A78AE v8.2 64-bit CPU 3MB L2 + 6MB L3
- DL Accelerator: 2x NVDLA V2.0
- Vision Accelerator: PVA v2.0
- Memory: 64GB 256-bit LPDDR5 2048GB/s
- Storage: 64GB eMMC 5.1
- Video Encode: 2x4K 60 | 4x4K 30 | 8x1080p 60 | 16x1080p 30(H.265)
- Video Decode: 1x8K 30 | 3x4K 60 | 7x4K 30 | 11x1080p 60 | 22x1080p 30(H.265)
- Power: 15W-60W
- Camera: 16-lane MIPICSI-2 connector
- PCle
The x16 PCIe slot supports up to x8 PCIe 4.0.
The M.2 Key M slot supports up to x4 PCIe 4.0.
The M.2 Key E slot supports up to x1 PCIe 4.0. - RJ45: Up to 10GbE
- USB
USB Type-C: 2x USB 3.2 Gen2
USB Type-A: 2x USB 3.2 Gen2,2x USB 3.2 Gen1
USB Micro-B: USB 2.0 - DisplayPort: DisplayPort 1.4a(+MST)
- microSD slot: UHS-1 cards up to SDR104 mode
- Other
40-pin header (12C,GPIO, SPI,CAN,I2S,UART, DMIC)
12-pin automation header
10-pin audio panel header
10-pin JTAG header
4-pin fan header
2-pin RTC battery backup connector
DC power jack
Power, Force Recovery, Reset buttons - Dimensions: 110mm x 110mm x 71.65mm (Height includes feet, carrier board, module and thermal solution)
DeepStream SDK
Building and deploying AI-based intelligent video analytics applications and services using DeepStream, which provides a multi-platform scalable framework with TLS security. It can be deployed at the edge and connected to any cloud.
Figure: DeepStream SDK
Simplifying workflow with the NGC Catalog
The GPU-optimized software hub for deep learning, machine learning, and high-performance computing accelerates deployment within development workflows.
Figure: Simplifying workflow with the NGC Catalog
NVIDIA Transfer Learning Toolkit
Accelerate AI training speed by over 10x and enable development like a zero-code professional.
Figure: NVIDIA Transfer Learning Toolkit
Embarking on the creation of an AI/ML system from scratch requires substantial investments of both time and resources, whereas transfer learning can efficiently extract learned features from an existing neural network model and apply them to a new model. The NVIDIA Transfer Learning Toolkit (TLT) is an innovative AI toolkit that significantly simplifies the complexity of AI/DL frameworks, enabling users to rapidly build production-quality trained models without coding.
