The most popular NVIDIA launched an artificial int

  • Detail

NVIDIA has launched an artificial intelligence computer with a power consumption of only 5 watts, which can create millions of intelligent systems

source: autor intelligent driving


on March 19, the GPU technology high pressure oil sealing was held in San Jose, California, the United States. At the sealing meeting using the oil film between the cylinder and the piston, NVIDIA announced today that it had launched Jetson nanotm, an artificial intelligence computer, Millions of intelligent systems can be created

nvidia Jetson nano developer kit is now officially available at $99. The Jetson nano module is priced at $129 (starting from 1000) and will be shipped in June. Both products will be sold through NVIDIA's major global distributors

this small but powerful CUDA XTM artificial intelligence computer provides 472 GFL for running modern artificial intelligence workloads. It is determined to control the raw materials in its own hands. It has the computing performance of OPS (billions of floating-point operations per second), and has high energy efficiency, but the power consumption is only 5 watts

nvidia founder and CEO huangrenxun announced at the GPU technology conference that two versions of Jetson nano were launched to provide developers, makers and technology enthusiasts with a developer kit priced at $99, and to provide enterprises that want to create edge systems for the mass market with a production ready module priced at $129

jetson nano supports high-resolution sensors, can process multiple sensors in parallel, and can run multiple modern neural networks on each sensor stream. It also supports many common AI frameworks, allowing developers to easily integrate their favorite models and frameworks into products

jetson nano has now joined the jetsontm series product lineup, which also includes the powerful Jetson AgX xaviertm for fully autonomous machines and Jetson TX2 for edge artificial intelligence. As an ideal choice for large enterprises, start-ups and researchers, the Jetson platform extends its scope of application to 30million makers, developers, inventors and students worldwide through Jetson nano

deepu Talla, vice president of NVIDIA and general manager of autonomous machine business unit, said: Jetson nano makes it easier for everyone to use artificial intelligence and is supported by the infrastructure and software adopted by the world's leading supercomputers. Bringing artificial intelligence into the maker movement has opened a new world of innovation and inspired people to create the next major product

jetson nano Developer Suite AI is far from being of maximum value in the maker community and education field, because general technologies often lack computing power and artificial intelligence software platforms

the $99 Jetson nano Developer Suite brings the power of modern artificial intelligence to a low-cost platform to support a new wave of innovation among makers, inventors, developers and students

they can build artificial intelligence projects that could not be realized in the past, and upgrade existing projects to a new level, including mobile robots and UAVs, digital assistants, automation equipment, etc

the suite provides out of the box support for full desktop Linux, is compatible with many common external devices and accessories, and provides readily available projects and tutorials to help makers quickly get started with artificial intelligence. In addition, the NVIDIA Jetson developer forum can also help people get answers to various technical questions

chris Anderson of DIY robots, DIY UAVs and the lonecode project of the Linux foundation said: the Jetson nano Developer Suite is eye-catching because it introduces advanced artificial intelligence into the DIY movement in a very easy-to-use way. We plan to bring this technology into our maker community because it is a powerful, interesting and affordable platform and a good way to teach in-depth learning and robotics to a wider audience

in the past, enterprises have been constrained by challenges such as size, power consumption, cost and artificial intelligence computing density. The Jetson nano module opens a new world for embedded applications such as video recorders, home robots and smart switches with complete analysis functions. It aims to save time for hardware design, testing and verification of complex, robust and energy-saving artificial intelligence systems, so as to shorten the overall development time and bring products to the market faster

the design is equipped with power management, clock, memory and fully accessible I/O. Because AI workloads are completely defined by software, enterprises can update performance and functionality even after system deployment

Sandeep Mehra, vice president and general manager of Cisco WebEx equipment, said: the mission of Cisco collaboration Conference (Cisco C cylinder piston is equipped with workbench ollaboration) is to connect all people and places to achieve immersive and immersive meetings. The cooperation with NVIDIA and the use of Jetson series products can achieve load control, which is the key to our success. Thanks to the advanced edge artificial intelligence processing capabilities of the Jetson platform, we have promoted new experiences to help people work better

to help customers easily transfer AI and machine learning workloads to the edge, NVIDIA has cooperated with Amazon AWS to support AWS IOT Greengrass to run Jetson nano and other Jetson driven devices perfectly

dirk didascalou, vice president of AWS IOT, said: our customers cover all walks of life, including energy management, industry, logistics, intelligent buildings and housing. Participants in all these industries are applying intelligence and computer vision to their applications to perform operations at the edge in near real time. AWS IOT Greengrass allows our customers to perform local reasoning on Jetson driven devices and return relevant data to the cloud to improve model training

nvidia cuda-x contains more than 40 acceleration libraries to support modern computing applications to benefit from the NVIDIA GPU accelerated computing platform. Jetpack sdktm, based on cuda-x, is a complete artificial intelligence software stack, including acceleration libraries for deep learning, computer vision, computer graphics and multimedia processing, and supports the entire Jetson series of products

jetpack includes the latest versions of CUDA, cudnn, tensorrttm and the full desktop Linux operating system. Jetson is compatible with NVIDIA artificial intelligence platform, which is a huge investment that NVIDIA has made for ten years and spent billions of dollars to promote Artificial Intelligence Computing Science

nvidia also created a reference platform to quickly start the development of artificial intelligence applications by minimizing the time spent on initial hardware assembly. NVIDIA jetbottm is a small mobile robot that can be built using off the shelf components and open source based on GitHub

jetson nano's main features include:

gpu:128 core GPU based on NVIDIA Maxwell architecture

cpu: quad core arm a57

Video: 4k@30 fps(h.264/h.265)/4k@60 FPS (h.264/h.265) encoding and decoding

camera: Mipi csi-2 dphy channel, 12 modules and 1 development kit

memory: 4 GB 64 bit lpddr4; 25.6 gb/s

connectivity: Gigabit Ethernet

operating system support: Linux for Tegra

module size: 70mm x 45mm

Developer Suite size: 100mm x 80mm

Copyright © 2011 JIN SHI