The graphics giant keeps making its play into the world of artificial intelligence
Nvidia has taken the wraps off Jetson TX2, a credit-card sized supercomputer aimed a powering artificial intelligence (AI) development and becoming the beating heart of embedded systems.
Following on from the Jetson TX1, the TX2 presents a 50mm x 87mm hardware development platform backed up with software from Nvidia, from which developers can use for prototyping products or to build intelligent systems around.
Nvidia sees the Jetson TX2 as a means to bring AI and other machine learning capabilities, including deep neural networks, to the edge of connected devices that make up smart networks and the Internet of Things (IoT)
“Jetson TX2 brings powerful AI capabilities at the edge, making possible a new class of intelligent machines,” said Deepu Talla, vice president and general manager of the Tegra business at NVIDIA. “These devices will enable intelligent video analytics that keep our cities smarter and safer, new kinds of robots that optimize manufacturing, and new collaboration that makes long-distance work more efficient.”
Jetson TX2 to the edge
The Jetson TX2 is not a radical departure form its predecessor, though its sports up-rated compute power, notably tapping into Nvidia’s latest Pascal architecture, and features beefier processors with the addition of a pair of Denver 2 CPUs to the array of quad-core ARM Cortex-A57 processors, and double the RAM and storage.
Other capabilities remain similar to TX1, though video encoding now hits 60 frames per second at 2160p resolutions and video decoding has 12 bit support for the H.265 and VP9 compression standards, which should make the Jetson TX2 capable of processing images effectively for running computer vision features in smart devices.
Jetson TX2 can also work in dual operating modes; the first allowing for maximum efficiency in power consumption, beating its predecessor by double the energy efficiency, while the second mode allows for maximum performance, with the new card beating its older sibling by double the performance.
Alongside the hardware specification, Jetson TX2 also supports JetPack 3.0, Nvidia’s software development kit for supporting the intergration of AI in to a variety of applications and product. The kit allows for both the training of AIs and putting them to use in the real-world through a process known as inference.
As more companies get involved in developing and deploying AIs and other machine-learning based systems, we can expect Nvidia to build out its Jetson portfolio and push its work in the field of AI.