NVIDIA has announced a range of new products from its embedded, automation, and robotics arm during its GPU Technology Conference keynote today — including the new Jetson Orin Nano module and IGX edge AI platforms.
"Computing is advancing at incredible speeds, the engine propelling this rocket is accelerated computing, and its fuel is AI [Artificial Intelligence]," NVIDIA founder and chief executive Jensen Huang claimed during his keynote speech at the GTC event today. "Today, we [announce] new chips, new advances to our platforms, and, for the very first time, new cloud services. These platforms propel new breakthroughs in AI, new applications of AI, and the next wave of AI for science and industry."
Among the new products and platforms announced at the event is the Jetson Orin Nano, a family of system-on-modules designed for edge AI workloads and delivering a claimed 80x performance boost over the company's prior generation of devices. Designed as lower-cost and lower-power alternatives to the top-end AGX Orin, the Orin Nano offers a choice of up to 20 or 40 trillion operations per second (TOPS) of sparse compute — cut in half for traditional dense workloads — with 4GB or 8GB of RAM, a six-core Arm Cortex-A78AE CPU, and either 512-core or 1024-core Ampere-architecture GPU with 16 or 32 Tensor cores.
“Over 1,000 customers and 150 partners have embraced Jetson AGX Orin since NVIDIA announced its availability just six months ago, and Orin Nano will significantly expand this adoption," claims Deepu Talla, vice president of embedded and edge computing at NVIDIA. "With an orders-of-magnitude increase in performance for millions of edge AI and ROS developers today, Jetson Orin is the ideal platform for virtually every kind of robotics deployment imaginable."
At the same time the company unveiled its IGX platform, powered by IGX Orin, its latest "AI supercomputer." While high performance and precision are key to the platform, they're not a sole focus. The company claims that the IGX platform combines software and hardware to deliver improved security and proactive functional safety in human-machine collaboration projects.
"As humans increasingly work with robots, industries are setting new functional safety standards for AI and computing," Huang claims. "NVIDIA IGX will help companies build the next generation of software-defined industrial and medical devices that can safely operate in the same environment as humans."
The same focus sits at the heart of NVIDIA's Isaac Nova Orin, an updated autonomous mobile robot reference platform built around the company's Jetson AGX Orin — either a single module in its base configurations, available with or without safety-certified sensors and programmable logic controller, or two collaborative modules, which handle vision-based functional safety in the top-end variant.
Another keynote announcement saw the unveiling of a cloud version of the Isaac Sim robotics simulation platform, giving developers the ability to access the platform through a new infrastructure-as-a-service offering dubbed Omniverse Cloud or Amazon's AWS RoboMaker — or to download and deploy it onto any public cloud platform themselves. "With these choices for accessing Isaac Sim in the cloud," claims NVIDIA's Gerard Andrews, "individuals and teams can develop, test and train AI-enabled robots at scale and in the workflow that fits their needs."
Not all of the announcements represent products and services available immediately, however. The Jetson Orin Nano modules will launch in January 2023, NVIDIA has confirmed, with pricing set at $199 for the 20-TOPS 4GB variant and $299 for the 40-TOPS 8GB model; those interested in the IGX platform can sign up now for notification when developer's kits go on-sale, with no word yet on pricing and a launch date.