NVIDIA Unveils its New Jetson Zavier NX, the World's Smallest AI Supercomputer

author: Eric Walz   

In the future, self-driving vehicles may roam the streets and small autonomous delivery robots may share the sidewalks with pedestrians while autonomous drones buzz overhead on their way to deliver the next package. All of these futuristic technologies will rely on sophisticated AI-powered hardware and software and companies like Nvidia are gearing up to meet the demand.

Nvidia revealed its brand new AI-powered processor today with enough computing power for the next generation AI-powered autonomous vehicles, delivery robots or drones. NVIDIA said the Jetson Xavier NX is the world's smallest AI supercomputer.

The Jetson Zavier NX is ultra-compact and about the size of a credit card, but it delivers server-class AI performance up to 21 TOPS for running modern workloads, including multiple neural networks in parallel and processing data from multiple high-resolution sensors simultaneously.  

The compact chip measures just 70x45 millimeters and has 384 CUDA cores, which is Nvidia's proprietary parallel computing programming model (Compute Unified Device Architecture), 48 tensor cores, plus two of Nvidia's open-source, Deep Learning Accelerators (NVDLAs).

The NX processor can be used in computationally demanding AI applications, such as training for image processing for autonomous cars or robots. 

"AI has become the enabling technology for modern robotics and embedded devices that will transform industries," said Deepu Talla, vice president and general manager of Edge Computing at NVIDIA. "Many of these devices, based on small form factors and lower power, were constrained from adding more AI features. Jetson Xavier NX lets our customers and partners dramatically increase AI capabilities without increasing the size or power consumption of the device."   

The compact Zavier NX is designed for embedded edge computing devices that demand high performance but are limited by size, weight, power constraints or cost, according to Nvidia.

In addition to self-driving cars, applications for the Zavier NX may include small commercial robots, drones, intelligent high-resolution sensors for factory logistics and production lines, portable AI-powered medical devices and other industrial IoT systems.   

The Jetson Zavier NX is compact.

The Zavier NX joins Nvidia's Zavier AI platform for autonomous machines, taking its place just under the top of the line Jetson AGX Xavier series and is backed by a full verification suite. The Zavier line of processors includes Nvidia DRIVE, a SoC for autonomous vehicles built from the ground up.

For companies already building embedded machines, Jetson Xavier NX runs on the same CUDA-X AI software architecture as the rest of the Jetson offerings, ensuring rapid time to market and lower development costs.  

Nvidia said its Jetson Xavier NX is already receiving strong support from the robotics and embedded devices ecosystem.  

In a statement, Lee Ritholz, Director & Chief Architect Applied Artificial Intelligence, Lockheed Martin wrote, "NVIDIA's embedded Jetson products have been accelerating the research, development, and deployment of embedded AI solutions on Lockheed Martin's platforms. With Jetson Xavier NX's exceptional performance, small form factor & low power, we will be able to do more processing in real-time at the edge than ever before."

When used in an autonomous vehicle, the Zavier NX can perform sensor fusion with data from up to 13 or more cameras in real time, running real time AI interference on computer vision algorithms for image classification.

Nivida designed the chips for high-performance needed to perform these tasks. In benchmark testing, the Xavier NX processor beat out Google's MobileNet-v1 computer vision models for TensorFlow as well as the deep neural network ResNet-50 v1.5 in performing image classification.

Jetson Xavier NX is also pin-compatible with Jetson Nano processor for small, low-power AI systems, allowing shared hardware designs and those developers with Nano carrier boards and systems to upgrade to Jetson Xavier NX. The NX also supports all major AI frameworks, including TensorFlow, PyTorch, MxNet, Caffe and others and runs all of Nvidia's CUDA-X AI models. 

NVIDIA CUDA-X is a collection of over 40 acceleration libraries that enable modern computing applications to benefit from NVIDIA's GPU-accelerated computing platform. 

Jetson Xavier NX is also supported by NVIDIA's JetPack software development kit (SDK), which is a complete AI software stack that can run complex AI networks, accelerated libraries for deep learning as well as computer vision and graphics.

Nvidia announced the Jetson Xavier NX will be available for $399 worldwide in March 2020.

Eric Walz

Originally hailing from New Jersey, Eric is a automotive & technology reporter covering the high-tech industry here in Silicon Valley. He has over 15 years of automotive experience and a bachelors degree in computer science. These skills, combined with technical writing and news reporting, allows him to fully understand and identify new and innovative technologies in the auto industry and beyond. He has worked at Uber on self-driving cars and as a technical writer, helping people to understand and work with technology.