March 17, 2026
CNBC Exclusive: Transcript: Nvidia Founder & CEO Jensen Huang Speaks with CNBC’s Jim Cramer on “Squawk on the Street” Today
WHEN: Today, Tuesday, March 17, 2026
WHERE: CNBC’s “Squawk on the Street”
Following is the unofficial transcript of a CNBC exclusive interview with Nvidia Founder & CEO Jensen Huang on CNBC’s “Squawk on the Street” (M-F, 9AM-11AM ET) today, Tuesday, March 17. More of the interview will air during “Mad Money” tonight at 6pm ET.
Following is a link to video on CNBC.com: https://www.cnbc.com/video/2026/03/17/nvidia-ceo-jensen-huang-our-growth-is-accelerating-at-a-larger-scale.html.
All references must be sourced to CNBC.
CARL QUINTANILLA: Joining us this morning in a CNBC exclusive is NVIDIA's founder and CEO, Jensen Huang, with our Jim Cramer. Hey, Jim.
JIM CRAMER: Thank you, Carl. Jensen, I do want to point out, someone said to me the stock's down 30 cents. Your stock's up 22,000 percent in the last decade. Should we care about the last half hour?
JENSEN HUANG: Well, I guess we all do. You know, our -- our growth is accelerating at a larger scale. That's surprising for people. And the end markets are really growing. You know, we've reached this point in A.I. called the inflection of inference. And so, we've reached the inference inflection. And so, you know, our growth is accelerating. We would love to love to see everybody enjoy it.
CRAMER: People got a little trapped by the 500 billion to a trillion. I like that over 2025, 2027. What trapped me was when you said, when we were talking about Michael Dell yesterday, that Jim, do you remember when I told you that ChatGPT could be big? And I said, well, yeah, I do. Ten years ago. And he goes, NemoClaw. He said, NemoClaw. This may be even bigger. I start my morning with NemoClaw. Could you explain to people why this may be what we're thinking about over the next five years and not something involving margins over the last five minutes?
HUANG: Well, first of all, I said -- I said at the keynote and I referenced a number I used previously because I want to give everybody a baseline. And so last year, I said that at this -- about this point, that we had 500 billion dollars of high confidence, visibility of Blackwell and Rubin. That's it. We obviously our company sells other things.
CRAMER: Right.
HUANG: We have, you know, CPUs, we have storage processors, we have networking. And so, we have all kinds of things that we sell in addition to that. We have $500 billion dollars' worth of visibility. And at this point, at this point, with another 21 more months to go to the end of 2027, we already have high confidence, high confidence visibility of $1 trillion dollar plus of Blackwell and Rubin, not anything else, just Blackwell and Rubin. And so, you know, we still have 21 more months of new orders coming in. We still have all of the other things that NVIDIA sells. And so, I just wanted to give everybody a baseline. The ecosystem needs to know directionally where NVIDIA is going. Now, of course, we plan with a supply chain. We help everybody with their capacity building. We prepare all the technology partners so that we can support a trillion dollars' worth of Blackwell and Rubin. And so, it's an incredible thing.
CRAMER: You bought Groq in December.
HUANG: Yeah.
CRAMER: What I heard yesterday, maybe my biggest takeaway was that it's already going to be making you not money, but a huge amount of money. And that's not part of the trillion.
HUANG: No. Yeah, we started production of Groq version three. It's a really, really special company. And -- and we're going to add Groq, not to every part. Vera Rubin is so incredible that it's really hard to add value to Vera Rubin. You know, that engine is so incredible. And so -- but there's some part of the workload, I think about 25 percent of the workload where we could add Groq to it. And so for 25 percent of the workload, we could increase our opportunity even more, and so—
CRAMER: People will be paying for this. One of the reasons why I think is that everyone's so worried about high bandwidth memory and the cost. I don't know correct if I'm wrong, but this kind of gets around that.
HUANG: We're the only company in the world --
CRAMER: The cost is going to be lower, your margins will be higher.
HUANG: Yeah. We're the only company in the world that now optimizes the entire data center across three different types of memories, HBM. And we have some excellent partners there, HBM4 coming, LPDDR5. The world's first low power memory used in data centers. We invented the whole idea. Everybody's going to use it in the future, but we're the only company in the world that optimizes around that today. And then third, we now have SRAM. So, we have three different types of memories that we could optimize the workload across. And you know, NVIDIA is really, really good at extreme co-design, where we optimize across all of these different technologies, building these A.I. factors.
CRAMER: And then at the same time, you talked about self-driving. Now, self-driving, that's 50 trillion. You're going to get your share of self-driving. Yesterday, I heard a dominant share. I did not know you would dominate in that market.
HUANG: We are going to be very, very large. You know, we've been working on self-driving for about 10 years now. Our strategy is not to build a self-driving car. Our strategy is to build a platform so that everybody can have self-driving cars. We partnered with Mercedes first. We're now on the road. It is the highest rated safety autonomous vehicle in the world today. And so, I'm very proud of that. We're now also in BYD, the largest electric car company in the world, Hyundai, Geely, and Nissan. Among the four -- consisting of all the five so far, that's 20 percent of the world's manufactured cars. And, yeah, we have GM and Toyota on top of that. And so, this is going to be a -- our strategy is to build, help Uber and help all these companies create a large fleet of autonomous vehicles.
CRAMER: Now, let's talk about the kind of robot that I didn't know I'd hear about, the agent robot. I saw yesterday for the first time that I have the opportunity to be smarter, to be able to create, to be able to do things I never thought because I've got a new buddy with me.
HUANG: Yeah.
CRAMER: It was -- it was viewed as soporific by the crowd. I thought it was explosive.
HUANG: Yeah, it was incredible. You know, there's a -- there's a guy named Peter Steinberger.
CRAMER: Yes, yes.
HUANG: And he created this thing called OpenClaw.
CRAMER: He's the first billion dollar whatever.
HUANG: He created this thing called OpenClaw. It reminds me of back in the old days when Linux first came out, or Kubernetes first came out, or HTML first came out for the internet. These open standards, at the time, came out modestly. In the case of OpenClaw, literally overnight, in just a few weeks, it was able to achieve the most popular open-source project in history, over, exceeding Linux, exceeding Kubernetes, exceeding every one of these open standards. And the reason for that is, it's this simple to download one line command and you create for yourself an agent, basically a digital robot. Instead of asking questions like what and how and where, which is what chatbots can answer, you ask it to do build, create, help, organize.
CRAMER: Cure.
HUANG: Cure. These are—
CRAMER: Build a house.
HUANG: That's right. Build, exactly. These are -- these are agents. These are basically an A.I. that can actually use tools and solve problems.
CRAMER: They're the fourth industrial revolution.
HUANG: That's -- that was our -- that's the dream always and OpenClaw made it possible and now it's completely open source.
CRAMER: I think you had to give everybody what they wanted yesterday. I think some took away an incremental notion. I took away an explosive notion and a wonderment notion. That's why I'm glad you're going to stay with me because we're going to talk a little bit more about not earnings per share, but a new world.
HUANG: Terrific.
CRAMER: A new world that NVIDIA's created—
HUANG: I love it.
CRAMER: And not just the world that made it so that you have so many millionaires.
HUANG: You get to see it right here at GTC.
CRAMER: Well, I'm very excited. I know -- one last thing I do want to just ask. Are we still doing dirty, dangerous, and dull robots so I don't have to do them?
HUANG: You're going to do, what?
CRAMER: The robots. They do dirty, dangerous, and dull things.
HUANG: Well, they -- they do things that -- that either we can't do, they do things that we don't want to do, and they can do things all the time.
CRAMER: Well, do they -- are they willing to clean toilets at Denny's?
HUANG: Nobody can clean toilets at Denny's better than I can.
CRAMER: That was your job.
HUANG: But that's, you know, that's a -- that's a gift.
CRAMER: Has there ever been a job that you weren't willing to do?
HUANG: Everybody's good at something. That's what I was good at.
CRAMER: So, I mean, were you happy with yesterday?
HUANG: Yeah. I'm proud -- I'm proud of all the ecosystem partners here. I'm proud of all the announcements that we made. We revolutionized A.I. factories again, introduced a level of performance and energy efficiency by another 10x, and then if you add Grok to that, 35 times, and so I'm very proud of that. I'm very proud of the fact that all of these ecosystem partners here are helping to build this future industrial revolution. The number of companies that are here is incredible. Every single industry is represented from healthcare to transportation to retail to manufacturing, right? It's just -- everybody's here, large companies—
CRAMER: Everyone is a client of Jensen Huang.
HUANG: It's incredible. It's incredible.
CRAMER: Well, we're going to have more with you, much more with you. Jensen's the Co-Founder and CEO of NVIDIA. You have to turn on “Mad Money” tonight to find out more, but it was explosive, not incremental, and that's why you must watch. Thank you, Jensen.
HUANG: Thank you, Jim.
For more information contact:
Stephanie Hirlemann
CNBC
