Cost-effective edge AI module Jetson AGX Orin 32GB is now available

Publisher:EEWorld资讯Latest update time:2022-08-04 Source: EEWORLDKeywords:Jetson  NVIDIA Reading articles on mobile phones Scan QR code
Read articles on your mobile phone anytime, anywhere

Bringing new AI and robotics applications and products to market, as well as supporting existing ones, is a major challenge for developers and businesses.


NVIDIA recently officially announced that the Jetson AGX Orin 32GB production-level module is now available.


Orin 32GB has a computing power close to 200TOPS. The specific parameters are as follows:


CPU: 12-core Arm Cortex-A78AE v8.2 64-bit processor with 3MB L2 + 6MB L3 cache

GPU / AI Accelerators:

2048 NVIDIA CUDA Cores and 64 Tensor Cores @ 1 GHz, powered by the NVIDIA Ampere architecture

DL Accelerator : 2x NVDLA v2.0

Vision Accelerator: PVA v2.0 (Programmable Vision Accelerator)

AI performance: 200 TOPS (INT8) @ 50W

Video encoding: 2x 4K60 | 4x 4K30 | 8x 1080p60 | 16x 1080p30 (H.265)

Video decoding: 1x 8K30 | 3x 4K60 | 6x 4K30 | 12x 1080p60 | 24x 1080p30 (H.265)

System memory: 32GB 256-bit LPDDR5 @ 204.8 GB/s

Storage: 64GB eMMC 5.1 flash

The 699-pin Molex Mirror Mezz connector supports:

Storage – Single-channel UFS

Display – 1x 8K60 Multi-mode DP 1.4a (+MST), eDP 1.4a, HDMI 2.1

camera

Up to 6 CSI cameras (16 via virtual channels)

16-lane MIPI CSI-2

D-PHY 1.2 (up to 40Gbps) | C-PHY 1.1 (up to 164Gbps)

Network – 1x GbE, 4x 10GbE

PCIe – 2x PCIe x8 (or 1x PCIe x8 + 2x PCIe x4), 1x PCIe x4, 2x PCIe x1 (PCIe Gen4)

USB – 3 USB 3.2, 4 USB 2.0

Low-speed IO – 4x UART, 3x SPI, 4x I2S, 8x I2C, 2x CAN, DMIC & DSPK, GPIO

Power: 15W, 30W or 50W

Dimensions: 100 x 87 mm


Nearly three dozen solution providers in the NVIDIA Partner Network around the world are offering products based on the new module, which delivers six times the performance of the previous generation.


With a broad range of products and services from Jetson partners, developers can build and deploy feature-rich Orin systems including action cameras, sensors, software, and connectors for edge AI, robotics, AIoT, and embedded applications.


The mass-produced systems with rich peripheral expansion capabilities enable customers to meet challenges in various industries from manufacturing, retail and construction to agriculture, logistics, medical care, smart cities, last-mile delivery, etc.


Helps quickly build more powerful AI products


Previously, developers and engineers had limited ability to handle multiple concurrent data streams in complex application environments. They faced stringent latency requirements, energy efficiency constraints, and various high-bandwidth wireless connectivity issues, while also needing the ability to flexibly manage software updates.


They also have to include multiple different types of chips in their designs to get the computing resources needed to handle diverse and growing amounts of data.


NVIDIA Jetson AGX Orin addresses all of these challenges.


The Jetson AGX Orin Developer Kit is capable of 275 trillion operations per second and supports multiple concurrent AI applications through the NVIDIA Ampere architecture GPU, next-generation deep learning and vision accelerators, high-speed I/O, and increased memory bandwidth.


With Jetson AGX Orin, customers can develop solutions using larger and more complex AI models to solve problems such as natural language understanding, 3D perception, and multi-sensor fusion.

image.png

The four Jetson Orin-based production modules released at GTC provide customers with a full range of server-grade AI performance. The Jetson AGX Orin 32GB module is available now, and the 64GB version will be available in November. Two Orin NX production modules will be launched later this year.


These production systems, powered by the NVIDIA Jetson software stack, enable thousands of businesses and millions of developers to build and deploy fully accelerated AI solutions on Jetson.


On top of the JetPack SDK, which includes the NVIDIA CUDA-X acceleration stack, Jetson Orin supports multiple NVIDIA platforms and frameworks, such as Isaac for robotics, DeepStream for computer vision, Riva for natural language understanding, the TAO Toolkit to accelerate model development using pretrained models, and Metropolis, an application framework, set of developer tools, and partner ecosystem that brings together visual data and AI to improve operational efficiency and safety across industries.


Customers can launch the next generation of edge AI and robotics applications faster by simulating any production-grade Jetson Orin-based module on the Jetson AGX Orin Developer Kit.


Expanding developer community and Jetson partner ecosystem


More than 1 million developers and 6,000 companies are building commercial products on the NVIDIA Jetson edge AI and robotics computing platform to create and deploy a wide range of autonomous machines and edge AI applications.


The growing Jetson partner ecosystem of more than 150 members offers broad support, including from companies specializing in AI software, hardware and application design services, cameras, sensors and peripherals, development tools, and development systems.


Approximately 30 partners offer commercial systems powered by the new Jetson AGX Orin module, with a variety of options to help support cutting-edge applications and accelerate time to market.


Developers who are selecting a complete development system will have their needs met with a range of options from AAEON, Auvidea, Connect Tech, Miwen, Pinli Technology, Ruitai New Era, and Tianjue Technology.


Allied Vision, Appropho, Basler, e-Con, Framos, Leopard Imaging, LIPS, RoboSense, Shenzhen Senyun Intelligence, Stereolabs, ThunderSoft, Unicore and Velodyne offer over 70 camera and sensor options that support challenging indoor/outdoor lighting environments and features such as LiDAR for mapping, localization and navigation of robots and autonomous machines.


Developers can get a full range of software support from Allxon, Cogniteam, Concurrent Realtime, Deci.AI, DriveU, Novauto, RidgeRun and Sequitur Labs, including device management, operating systems (Yocto and Realtime OS), AI software and tool suites.


Developers who need connectivity options such as WiFi 6/6E, LTE, and 5G can check out offerings from Telit, Quectel, Infineon, and Silex.


The new NVIDIA Jetson AGX Orin 32GB production-grade module is available in China through the following distribution channels: China Electronics Port Electronic Mall, Synnex International and Arrow Electronics.


Keywords:Jetson  NVIDIA Reference address:Cost-effective edge AI module Jetson AGX Orin 32GB is now available

Previous article:In embedded development, what is the difference between embedded hardware and software?
Next article:Raspberry Pi 4 now supports Vulkan 1.2

Recommended ReadingLatest update time:2024-11-16 11:40

Ready to use: NVIDIA and VMware deliver AI at scale for the enterprise
NVIDIA and VMware bring AI to the enterprise at scale NVIDIA AI Enterprise and VMware vSphere with Tanzu Simplify Enterprise AI Development and Application Management   NVIDIA and VMware have reached another milestone in their collaboration by jointly developing an AI-ready enterprise platform that brings the wo
[Home Electronics]
Ready to use: NVIDIA and VMware deliver AI at scale for the enterprise
Nvidia enters medical industry with AI tools to the American College of Radiology
AI can be applied in more areas than just games. Recently, NVIDIA has opened up a new path for AI and reached a partnership with the American College of Radiology (ACR), committed to providing medical personnel with faster and more effective image processing solutions.   It is understood that the technology is still
[Medical Electronics]
Nvidia enters medical industry with AI tools to the American College of Radiology
News says Nvidia GeForce RTX 4060 Ti and RTX 4060 will be launched in May
According to Wccftech, NVIDIA will launch the GeForce RTX 4060 Ti and RTX 4060 GPU for the mainstream market in May this year. According to reports, the RTX 4060Ti will use a 160W AD106 core, while the regular version will use a 115W AD107 core. NVIDIA GeForce RTX 4060 Ti specs "rumored": The RTX 4060 Ti is expecte
[Home Electronics]
News says Nvidia GeForce RTX 4060 Ti and RTX 4060 will be launched in May
Nvidia RTX 40 series graphics card production is reported to be reduced by 50%, Huang Renxun: Blackwell samples will be sent this week
On July 30, Huang Renxun attended the SIGGRAPH 2024 summit held in Denver, USA yesterday (July 29), and said that Blackwell architecture samples will be sent to customers starting this week. The Economic Daily, a news source, also revealed yesterday that Nvidia has significantly cut the supply of consumer graph
[Home Electronics]
Nvidia's 40 billion pounds attack on ARM has entered the negotiation stage
According to foreign media reports, Nvidia will acquire ARM for 40 billion pounds (52.4 billion U.S. dollars at the current exchange rate). Currently, Nvidia and ARM have entered the exclusive negotiation stage, and the overall acquisition will be completed by the end of this summer.   Gu Wenjun, chief analyst at S
[Embedded]
Nvidia's 40 billion pounds attack on ARM has entered the negotiation stage
NVIDIA and Hugging Face will connect millions of developers to generative AI supercomputing
The integration of NVIDIA DGX Cloud into the Hugging Face platform will accelerate the training and tuning of large language models (LLM), simplifying model customization in nearly every industry   LOS ANGELES — SIGGRAPH — August 8, 2023 — NVIDIA and Hugging Face announced a partnership to provide generative AI su
[Industrial Control]
NVIDIA and Hugging Face will connect millions of developers to generative AI supercomputing
Artificial intelligence chip business drives NVIDIA's performance surge
On August 23, local time, the American chip giant Nvidia announced its second quarter financial report as of July 30. The report shows that due to the strong demand for artificial intelligence chips, the company's net profit in the second quarter reached a record high of US$6.7 billion, a surge of 422% year-on-year; r
[Semiconductor design/manufacturing]
From generative AI to physical AI, Huang Renxun made a new judgment
At the 2024 GTC conference, NVIDIA CEO Huang Renxun demonstrated 9 humanoid robots in one go. On Sunday, he once again emphasized robots and AI in his Computex 2024 speech. Today, more and more customers are choosing the Isaac platform and announcing more details of the cooperation and products, which a
[Industrial Control]
From generative AI to physical AI, Huang Renxun made a new judgment
Latest Embedded Articles
Change More Related Popular Components

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号