NVIDIA introduced new developer platforms to power the next generation robotics and autonomous vehicles. NVIDIA founder and CEO Jensen Huang unveiled the Jetson AGX Xavier (robotics) and DRIVE AGX Xavier (self-driving cars) at the company’s GTC Japan event.
The Jetson AGX Xavier includes a small computer module capable of up to 32 TOPs (trillion operations per second) and multiple operating modes at 10W, 15W and 30W. NVIDIA says the Jetson AGX Xavier has greater than 10x the energy efficiency of its predecessor.
Jetson AGX Xavier also supports a full software stack for developing and deploying AI applications through the NVIDIA JetPack SDK, which includes the latest versions of CUDA, cuDNN and TensorRT. The NVIDIA Jetson AGX Xavier developer kit is available now for $2,499. Members of NVIDIA’s Developer Program are eligible to receive their first kit for $1,299.
“Japan is driving the world of robotics in numerous industries – from factory automation to construction to manufacturing,” says Deepu Talla, vice president and general manager of Autonomous Machines at NVIDIA. “The country’s most important companies in this space are adopting Jetson AGX Xavier to usher in a new era of robotics.
Some of those companies already adopting Jetson AGX Xavier include FANUC, Komatsu, Musashi Seimitsu and Kawada Technologies. Yamaha is also using Jetson AGX Xavier to power its upcoming autonomous machines, which includes unmanned agriculture vehicles, last-mile autonomous vehicles and autonomous marine products.
Yamaha says these unmanned agriculture robots and drones are vital because of Japan’s aging population, where the average worker is 66.6 years old and the workforce has declined more than 30 percent from 2010, according to the Japan Ministry of Agriculture. The agriculture robots and drones will help spray farmland, pick fruits, and transport harvested crops and equipment. Yamaha expects to begin testing these devices in 2019, with the goal of a public launch shortly afterward.
Jetson AGX Xavier’s 16GB of 256-bit LPDDR4x memory provides 137GB/s of bandwidth are designed for AI-powered vision and multimedia applications processing high-bandwidth sensors like HD cameras and LIDARs with low latency. It can encode up to four 4Kp60 streams and decode two 8Kp30 with hardware video codecs. High-bandwidth I/O includes 16x MIPI CSI-2 lanes to connect six sensors directly or up to 36 virtual channels, 8x SLVS-EC lanes, and five PCIe Gen 4 controllers (1×8, 1×4, 1×2, 2×1) with support for root complex and endpoint modes.
The Jetson AGX Xavier embedded compute module measures 100x87mm and utilizes a high density 699-pin board-to-board connector for breaking out the I/O signals to carrier boards and user-defined systems.
NVIDIA DRIVE AGX Xavier
For autonomous vehicles, the NVIDIA DRIVE AGX Xavier Developer Kit comes with a DRIVE AGX Xavier car computer, along with a vehicle harness to connect the platform to the car, international power supply, camera sensor and other accessories. The platform runs the NVIDIA DRIVE Software 1.0 release, which incorporates a range of operations necessary for self-driving, including data collection, obstacle and path perception, advanced driver monitoring and in-vehicle visualization.
The release includes DriveWorks modules for a sensor abstraction layer, as well as computer vision and image processing libraries that streamline autonomous driving development. The platform is open, enabling companies to build their own applications on the base software, and enhance it via over-the-air-updates.
The NVIDIA DRIVE AGX Xavier Developer Kit will be available on October 1. No price has been given.
Subhas Hati says
I would like to know whether the software for AGX Xavier is included in the price. I find that it has CPUs. So, is it a full board computer or an extra PC/Laptop is needed.