This week’s news stories focus on NVIDIA’s annual GPU 2024 conference and the major impact NVIDIA’s announcements will have on the future of robots, cobots and AMR.
NVIDIA’s breakthrough moment for robotics
A brave new world of artificial intelligence for robots, cobots and AMR
One corporate event for 2024 is of huge significance to all of robotics: Santa Clara-based NVIDIA’s annual GPU extravaganza. What’s new in robotics this week? will spend most of their digital ink covering the event.
Autonomy now means “intelligence”. If you’re a robot, cobot, or AMR, you’re in for a treat unless you can get some smarts by converging with AI ASAP. NVIDIA boss and co-founder Jensen Huang just opened a few doors for you at the 16th GPU Technology Conference (GTC) (March 17-21) and then graciously invited you to jump in to get a little better.
It was Huang’s version of March Madness, and during NVIDIA’s annual keynote, he evoked thrills, chills and thrills in a three-pointer tone. For his online audience of 250,000 (10,000 in person), he didn’t disappoint: Huang presented a vision of the future of computational and generative artificial intelligence and how it just kicked the door off Moore’s Law… how it keeps accelerating.
The importance of bots to AI and vice versa cannot be emphasized enough. Without bots, there are no AI GPUs; robots make AI hardware where generative AI does its magic.
Here’s a video showing how GPUs (which stands for graphics processing unit) are built. No bots, no GPUs. Reaching the tipping points of accelerated computing and generative artificial intelligence cannot be achieved without robot-built GPUs.
Without GPUs, automating advanced storage and manufacturing through digital twins is impossible. This video shows how closely related each other is. As Huang says in the video: “In the future, everything that moves will be robotic.” Think about that for a moment. The place of robotics in the future of logistics, manufacturing and society is central to everything.
“Generative AI is the defining technology of our time,” Huang said. And NVIDIA’s latest GPU, the Blackwell, with its 208 billion transistors, claims to power a new industrial revolution. Huang also showed the role of digital twins, which can simulate, test, define and redefine “large-scale artificial intelligence in real-time before introducing it into an industrial environment.” Saving money for sure! Using video from a warehouse above, he demonstrated digital twins acting as “air traffic control” to monitor autonomous machines and autonomous people below. Best of all, you only need a browser to run everything.
For example, an SME could use a cloud-based digital twin to pre-plan a space for a robot to use in a factory or warehouse, customize it exactly to their needs, and then run test runs of all aspects, including raw materials. , robots, conveyor systems and personnel before you buy a single robot or cobot.
Amazingly, NVIDIA made nearly forty “standalone announcements” at GTC, including the showstopper: the 208 billion transistor Blackwell. “the world’s most powerful chip”, which contains 208 billion transistors. Called Blackwell, the chip is aimed at applications such as deep learning, engineering simulation and AI.
In a demonstration of brute force, NVIDIA’s Metropolis visual artificial intelligence platform created a map of worker activity across the warehouse, “combining data from 100 simulated ceiling cameras with multi-camera surveillance.” Maps created by Metropolis will help optimize AMR routes.
Teradyne Robotics
Some robot/cobot vendors, no doubt seeing the inevitability, benefits and also the huge potential of GPUs, seem to have been working with NVIDIA for some time. North Reading, MA-based Teradyne, owner of Danish cobot developer Universal Robots (UR) and AMR MiR specialist, was one such vendor toying with NVIDIA’s GPU offerings.
As a lifelong Bostonian, I love that the Teradyne circle is bringing its two robotics entities under its corporate wing as Teradyne Robotics.
The move may also bode well for future UR robotics, which will gain an insider opportunity to participate in Teradyne’s semiconductor business. Thanks to Teradyne’s partnership with Italy’s TechnoProbe, we can well see cobots and probe cards working together.
At GTC, Teradyne Universal Robots demonstrated an autonomous AI-powered inspection system using the Nvidia Jetson AGX Orin edge AI computer that makes robot path planning 50-80 times faster than today’s systems. “MiR uses the same module in the new AMR pallet jack, which uses 3D vision to identify, pick and deliver pallets with ‘unprecedented accuracy’, even in dynamic and complex environments.”
NVIDIA also announced “a collection of pre-trained models, libraries and reference hardware for robot developers. Its Isaac Manipulator platform offers modular AI capabilities for robotic arms as well as GPU-accelerated libraries. It can speed up path planning by 80 times and allow developers to automate more robot tasks. Early adopters include Yaskawa, Universal Robots, PickNik Robotics, Ready Robotics and Franka Robotics.
Although NVIDIA seems to own and lead the world in GPUs, NVIDIA’s stock price has fallen a bit, according to MarketWatch.
“Shares of NVIDIA Corp. NVDA was down 2.50% at $902.50 on Wednesday in what turned out to be generally favorable trading for the stock market, with the S&P 500 SPX up 0.86% to 5,248.49 and the Dow Jones Industrial Average DJIA up 1.22% to 39,760.08.
It just goes to show that you can’t please everyone.
Undeniably, regardless of its shared performance, NVIDIA has taken robotics to new heights. Here’s a final video that offers a glimpse into the future of robots and GPUs. Enjoy.
