Nvidia’s CEO sees AI revolution transforming energy grid
Artificial Intelligence will require more energy in the future but will also help energy companies become more efficient in delivering those needs, Jensen Huang, CEO of Nvidia said Tuesday
Speaking during the opening session of the Edison Electric Institute’s annual meeting in Las Vegas, Huang said while data centers, artificial intelligence and accelerated computing will require larger amounts of power, the benefits of those systems will allow electric companies to provide energy with less waste and more efficiency.
Nvidia, the global leader in accelerated computing and generative AI, became the world’s most valuable company on Tuesday with a stock market value that topped $3.34 trillion.
In conversation with Huang, Pedro Pizarro, EEI chair and president and CEO of Edison International, said the tech industry and energy industry are fundamentally linked, and that growth in the tech industry has increased electricity demands as computing changes.
Pizarro asked Huang how the two industries could work together in the future.
“We are seeing electricity demand, which had been fairly flat for many years, now really taking off,” Pizarro said. “This has given you a unique perspective about the intersection between the energy industry and the technology industry.”
While the age of artificial intelligence and advanced computing has increased demand, Huang said, it will also work to increase the efficiency in how the industry uses that energy and in how utilities can deliver energy.
Huang said that up until the AI revolution, all computing was general computing and was akin to using a general-purpose tool to solve all problems. And using a more intelligent solution is energy efficient, but also leads to increased demand for that solution.
“The first thing that we were doing is that we’re accelerating every application we can,” Huang said. “And when we accelerated an application, not only do we speed it up, we reduce the cost of computing, we reduce the energy consumed by it… that would suggest, as a result, the amount of energy we use would go down. But in fact, it doesn’t. And the reason for that is when you make something cost effective, if that’s something that is valuable, then the consumption of it will go up.”
AI will change the face of the energy industry, he said, not just by how much demand will increase, but in where companies will put their grids.
“One of the things that you should keep in mind is that today’s energy has built around where society is, where population is, where the grid is,” Huang said. “In the future, you probably generate electricity and generate energy near grids and not near grids. That’s something to consider as you do this because you have to keep in mind that AI doesn’t care where it goes to school. It doesn’t have to be near population… We’re still going to require a lot of electricity, a lot of energy, near population, but you can also create quite productive energy away from the grid.”
Partnering with AI to better provide those energy needs is an important step in the future, Pizarro said.
“I know a number of us are engaged with your team in areas like smart meter design, where should the next generation of smart meters scale, more broadly, a number of us are deploying AI across our companies and all sorts of operational sort of functions,” Pizarro said. “I agree with the thesis that the more we can use these tools to be more efficient, the better it is, in terms of our overall mission, to provide power reliably, safely, affordably and importantly, making sure it’s clean.”
Huang said he envisioned several ways the two industries can collaborate.
“You could use AI inside the company to be more productive, eliminate waste, go faster, do more with less, which is what fundamentally intelligence does,” he said. “However, what is likely the greatest return or the greatest impact will likely be applying AI in the delivery of the energy itself, the provisioning of the energy grid.”
Huang said he sees electricity providers operating something akin to an app store where consumers who generate excess energy could provide it to consumers who need more energy.
“In a lot of ways, all of your power grid today, which is really just a distribution network, is going to become a smart network like the internet,” he said. “That’s kind of like Uber, kind of like an app store. And if that’s the case, then how do you make that happen? Well, the way to make that happen is smart meters… This is a great opportunity for all of you, your business model could really be transformed.”
AI and accelerated computing will also help the industry when it comes to cybersecurity, Huang said. As criminals and hackers become more sophisticated, it will be even more important for utilities and co-ops to use smarter tools to thwart attacks.
“Essentially, you’re having computers attacking computers, and when you have computers attacking computers, you’d like to be the faster computer, “ Huang said. “Attackers are increasingly using accelerated computing and generative AI as a way to penetrate the system’s defenses. We’re seeing an increasing adoption of accelerated computing, because it’s so fast… that you have the benefit of doing two things, you could respond faster and you could apply much more sophisticated defenses.”
But Huang pushed back on the idea that AI would eliminate jobs, saying instead that increased productivity across industries will lead to more jobs, not fewer. But, he said, the jobs may be different from the jobs we have today.
“Most companies or most industries I know, when they become more productive.. they become more profitable. And when they become more profitable, they tend to hire more people… The world’s GDP is the largest has ever been. And the reason for that is because we’re more productive than we’ve ever been endless because we also have more ideas than we’ve ever had. So long as the world has ideas, if we were more productive, we will have increased earnings to fuel to fund those ideas, which drives up the size of the industries… And those ideas are going to create jobs, however, the jobs will obviously change.”
