Nvidia CEO Jensen Huang during a media briefing from Computex in Taiwan this week called enterprise AI “the largest opportunity of all,” described his metaverse vision and expressed patient confidence about Nvidia’s chances of closing its planned $40 billion acquisition of Arm.
In a typically wide-ranging and mind-expanding discourse Wednesday morning in Taiwan, Huang suggested that advancements in deep learning and natural language understanding are driving AI to new levels. “...Using machines, you could write software that no humans possibly can,” he said, adding, “There's a whole bunch of different algorithms that are associated with deep learning, but there's no question that we can now write software that we would not have been able to write before. And so we can automate a whole bunch of things that we never thought were going to be possible in our generation.”
But, Jensen said the ability to tap into the most advanced aspects of AI has been limited to the largest internet and cloud companies. Nvidia, through hardware announcements it made at its GTC event in April and software announcements it made this week at Computex, is looking to change that, broadening the market for AI and making it more accessible to more enterprises.
For example, the Nvidia Certified Systems program, which the company said this week it is expanding, makes it easier for more enterprises to use the Nvidia AI Enterprise software stack to gain access to AI tools and pre-trained AI models that they can leverage for their own industrial use cases.
“A pre-trained AI model is kind of like a new college grad, they got a bunch of education, they're pre-trained, but you still have to adapt them into your job, into your profession, into your industry,” Huang said.
He added that having access to the software on high-volume servers all over the world is “like having an AI cloud, but it's in your company. And it comes with a bunch of tools and capabilities for you to be able to adapt. Now, how would you use it? Healthcare would use it for image recognition for radiology, for example, retail will use it for automatic checkout, warehouses and logistics, moving of products and things like that, tracking inventory automatically. And cities would use these to monitor traffic. Airports would use it in case somebody lost their baggage, it would instantly find it. So there are all kinds of different applications for AI in enterprises. And I expect enterprise AI, what some people call the industrial edge, will be the largest opportunity of all…”
Entering the metaverse
Regarding the emergence of the metaverse, that is to say the notion of many virtual worlds connected to the physical world, Huang said “we are right on the cusp of it,” adding that “it has real design, has real economy. And you have a real avatar, and that avatar belongs to you and is you. It could be an avatar. It could be a photo. It could be a character. And in these metaverses, you'll spend time with your friends.”
As he spoke directly to a large video screen featuring the faces of dozens of media representatives attending the online briefing, said, “In the future, we will be in a metaverse right now. It will be a communications metaverse. It won't be flat, it'll be 3D. And we'll be able to almost feel like we're there with each other. It's how we do time travel. It's how we travel to far places at the speed of light. That's how we go backwards in time. That's how we go forward in time. Going forward in time is to simulate the future.”
Huang said there will be video game metaverses for games like Fortnite and World of Warcraft, but that the concept also will extend to business and industry. “We're going to see a metaverse overlay, if you will, into our physical world, the world of industry.” For example, he explained that the building he was standing in at the moment, a new Nvidia facility in Taiwan that will serve as the workplace for 3,500 people, was designed using virtual reality technology and that it exists in virtual reality as a digital twin.
At the same time, the physical building itself was designed to accommodate robots, for many reasons, but one being the potential for employees to be virtually present through robots. “You could upload yourself into a robot and sit at your desk with your virtual reality headset or AR headset and you can roam around the campus as if you're there,” he said.
Closing the Arm deal
Back in the much more mundane physical world, Nvidia is still working through the paces of acquiring Arm. Many reports have questioned if the deal will survive the international regulatory approval process intact, but Huang said he was “confident” that it will, and expressed a sense of patience with a process that he expects to last 18 months before regulators in the U.S., U.K. and China all chime in.
“That's the typical journey. Mellanox took about 18 months or close,” he said. “I expect this one to take about 18 months. That makes it early next year, later this year, early next year. I'm confident about the transaction the regulators are looking for. Is this good for competition? Is it pro-competitive and brings innovation to the market? Does it give customers more choice? Does it give customers more offerings and more choice?”
Huang argued that the Arm acquisition is good for ongoing competition and customer choice because it will massively scale Arm’s R&D capabilities and resources to allow it to develop more advanced products more quickly and to reach many more customers.
“If we combine NVIDIA and Arm, Arm's R&D scale will be much larger," he said. "As you know, Arm is a big company, Arm is not a small company, but NVIDIA is much bigger. Our R&D budget is many, many times larger than Arm. Our combination will give them more R&D scale. It would give them technology that they don't have the ability to build themselves or the scale to build themselves like all of the artificial intelligence expertise that we have. We can bring those capabilities, through Arm, to its market and as a result of that, we will offer Arm customers a more technology choice, better technology, more advanced technology. And that, ultimately, is great for competition because it allows Arm's licensees to create even better products, more vibrant products and better leading edge technology, which… would give the end market more choice.”
He continued, “You could see that on first principles, because our companies are completely complimentary--they build CPUs, we build GPUs and DPUs. They don't build GPUs, really... I think our companies are complimentary and so we'll bring, by nature, innovations that come as result of companies that come together that offer complimentary things. So it's like ketchup and mustard coming together. It's good for innovation.”
RELATED: Nvidia CEO Jensen Huang and the future of… everything