FOX BUSINESS ALERT, THE THOUSAND-POINT MARKET SWING WHERE THE EARLY RALLY WAS COMPLETELY TORCHED HAS INVESTORS' HEADS SPINNING. BUT NVIDIA'S JENSEN HUANG IS LOOKING SKYWARD AFTER HIS COMPANY'S THIRD QUARTER EARNINGS SMASHED EXPECTATIONS. LOOK, LET'S BE CLEAR, NVIDIA HAS NOT ESCAPED BEING BITTEN BY THE JAWS OF TODAY'S VOLATILE. IT'S AT 181 RIGHT NOW, AT ITS HIGHS, 196, IT'S NOW DOWN 2.6%. THE A.I. CHIP KING'S REVENUE SURGE SHOWS ITS BUSINESS IS CHARGING TOWARD. IT IS THE PREMIER SELLER OF THE BRAINS RUNNING THE A.I. WORLD'S AMBITIONS, AND THAT'S WITHOUT ACCESS TO CHINA'S $50 BILLION CHIP MARKET. I SPOKE WITH THE LEADER OF THE $4.5 TRILLION COMPANY FROM WASHINGTON, D.C. WHERE HE WAS CONSPICUOUSLY MISSING HIS LEATHER JACKET. CONSIDERING HIS BLOWOUT THIRD QUARTER REPORT, I ASKED JENSEN HOW THE CURRENT QUARTER LOOKS, WHEN AND HOW SMOOTHLY WILL THE NEW RUBIN CHIP ROLL OUT, AND WHETHER HE'S WORRIED ABOUT AN A.I. BUBBLE. JENSEN, THANKS FOR JOINING US. $51.2 BILLION OF YOUR $57 BILLION REVENUE NUMBER CAME FROM DATA CENTER CHIP SALES. THAT IS MONSTROUS. I'M SO CURIOUS TO KNOW, WHAT'S THE LARGEST SINGLE ORDER TO DATE THAT YOU'VE GOTTEN?
Oh, my gosh, I don't know. They're all in the billions. You know, they all have to be in the billions in order to add up to 51. So I don't know exactly which one. But one of the things that is really great about NVIDIA's architecture is that we're so diversified. You know, we run — A.I. is not just one thing, it's not just one language model or one chatbot. A.I. is a system of models, and you can use it for, of course, chatbots, but you can also use it for physics and biology and chemistry and, of course, robotics and self-driving cars and even within language models there's so many different architectures. NVIDIA's architecture runs literally everything, and yesterday we announced a partnership with Anthropic, and now we're the only architecture in the world that runs every single premier A.I. model as well as every single open source model and, of course, every single version, modality of model. So our versatility makes it possible for us to have a lot of different customers, and we can bring them a lot of other customers that ultimately use their cloud capabilities.
Yeah. It feels like all roads are leading to NVIDIA. How are orders for your next generation chip, the Rubin, trending, and are you still on track to ship them at the end of 2026?
I announced at GTC that starting this year, because we started shipping Blackwells, over the course of this year through the end of next year we have visibility of half a trillion dollars' worth of Blackwell and Rubin. And we'll start shipping Rubin in Q3 time frame of next year. The silicon's back, it is working wonderfully. Engineers across NVIDIA are bringing up the system, doing it beautifully, and we're on track for volume production in the Q3 time frame of next year. In the meantime, I announced that we have half a trillion worth of visibility, but in the last 30 days we announced some major new wins that were not included in that half a trillion dollars. So we have, we're adding, of course, the Anthropic opportunity, we're adding Elon's, and A.I. now being deployed in Saudi Arabia. We announced a partnership with Humane, and those GPUs have been licensed, and they're going to be deploying in Saudi Arabia. And we also announced that AWS is going to also build 150,000-GPU A.I. infrastructure in Saudi Arabia. All of those are net new, and so orders are still coming in. It's going to be a pretty busy year for us.
It sounds like then the current quarter is going to be crazy good. Can I say that? What do you think?
Well, we guided to a much larger quarter next quarter, and so the guidance that we provided is crazy good, I would agree with that. But, you know, we're in the beginning of a very long-term buildout of the fundamental infrastructure of humanity which is computing. As you know, we've reinvented computing for the first time in 60, 70 years. So all of the computers that have been installed around the world are being modernized to accelerated computing and NVIDIA's GPUs and to artificial intelligence. So this buildout is going to last us many years to come.
It was exactly a year ago, and this is why I ask about Rubin orders, etc., because it was a year ago that you and I spoke about the problem a lot of companies wish they had. Blackwell orders were pouring in so fast. Demand outweighed supply. What specifically have you changed to prevent any bottlenecks when Rubin rolls off the assembly line?
We're not going to be caught flat-footed too many times, and we have a much better understanding of the scale and the breadth, the diversity of our customer base. We've done a really good job planning our supply chain. We now have the largest supply chain of any company in the world. We work with TSMC, of course, manufacturing in Taiwan and also, very proudly, for the first time here in the United States. We work with every single memory vendor now. We have three gigantic sources of really complicated memory, and every system maker in the world are partnering with us. And so we've been planning for the Rubin launch, and we've done a really good job getting set up for that. It's going to be a really big year next year.
In September you made a minor investment in Intel. Would you consider moving some of your physical manufacturing to Intel's foundry business? Just to spread your bets?
