November 11, 2026
WHEN: Today, Wednesday, November 11, 2026
WHERE: CNBC’s “Squawk on the Street”
Following is the unofficial transcript of a CNBC interview with CoreWeave Co-Founder, Chairman & CEO Mike Intrator on CNBC’s “Squawk on the Street” (M-F, 9AM-12PM ET) today, Wednesday, November 11 where he discussed the company’s Q3 revenue surges, the demand for computing infrastructure, views on Nvidia, and more. Following is video of the interview on CNBC.com: https://www.cnbc.com/video/2026/11/11/coreweave-ceo-mike-intrator-the-demand-for-computing-infrastructure-remains-overwhelming.html.
All references must be sourced to CNBC.
JIM CRAMER: Michael Intrator, welcome back to the show.
MICHAEL INTRATOR: Thank you for having me, guys. Great to be here.
CRAMER: Okay. I don't know where to start because the pluses were actually so good. Your revenue was amazing. The adjusted loss was much less than I thought. Margins improved a quarter ahead of the plan. But I think maybe if you can just start by saying the demand and how outsized it is, because there were some analysts who still seem to think that this is the tail end of this revolution.
INTRATOR: Yes. That's, that doesn't really actually track with the facts on the ground. The demand for this infrastructure, and I've said this to you guys now for years, outpaces the world's capacity to deliver infrastructure. You know, and that is the case. It's, you know, I think I said that we are 10x oversubscribed to the existing capacity that we have left in 2027, and we are multiples overdescribed, oversubscribed on the capacity we have out in 2028. So, I heard those questions, too, and I would like to, you know, reiterate what you're hearing across the market, which is the demand for the infrastructure is and remains overwhelming.
CRAMER: There is a story that we've been believing since the so-called blue wave, which is that basically we're not going to have a lot of new data centers because of what's happening in the United States. Your international exposure makes me think that maybe we're being too parochial about our own country. There is worldwide demand, and it's not, it's growing. It's insatiable.
INTRATOR: Yes, that's certainly true. There is a massive groundswell of demand around the world for computing infrastructure. But, you know, taking a moment and talking about the U.S., right, you know, we don't see any impact on our ability to deliver the data centers within the U.S. that are required for us to deliver on our backlog. You know, in Texas, governor-elect, you know, said that, you know, facilities that are permitted and under constructions will continue. And, you know, we've been really, really thoughtful about the broad distribution.
CRAMER: And you sent her a letter which just said basically listen, we're good citizens. I like that. Now, I think it's very important that we get to the fact that a lot of us were grading your company by looking at public debt and thinking 15, 16. Well, that is very close to where I regard it as a danger zone. You pretty much cleared that up by doing two things. One, the H100s are worth a lot more than you think when you get them, but also you indicated there's another kind of debt to the company that actually is basically trading in part.
INTRATOR: That is correct. Our asset level debt is basically been rock solid through this entire thing. There is an enormous demand for that product. We're about to launch another asset level debt in the next week or two that has already been rated high investment grade. You know, we suspect that will be underwritten in the market at a 7 to a 7.5 percent interest is what we expect, and that will once again bring down the average cost, average weighted cost of capital for CoreWeave, and that's really, really exciting. You know, it is been a priority for the company to monitor that. And, you know, the high-yield debt is just one corner of how CoreWeave accesses capital, and you saw that in our earnings when I said, okay, high-yield debt is getting expensive. We'll go to prepayments, which we have done in the past, and we raised $9 billion worth of prepayments and deposits. And, you know, that's just another way for us to go about financing the business to ensure that we have the capital at reasonable costs to be able to execute on our roadmap.
DAVID FABER: Right. I mean, a key focus for investors is adjusted operating income and interest expense. And, I mean, I am still looking. I think in the fourth quarter, interest expense is still going to outpace your adjusted operating income. When do those two actually change trajectories, and so your interest expense is actually going to be less than your operating income, adjusted?
INTRATOR: So, our interest expense went up by 20 percent. Our adjusted operating income went up by 300 percent, right? If you look at the trajectory, it is clear that our interest expense, while growing, as it should, because we are delivering massive scale of infrastructure. You know, we have 1.7 gigawatts in the ground. We will get to 1.85 gigawatts by the end of the year. That is massive amounts of infrastructure. And so, yes, the debt is growing and the interest expense associated with it, but it is not growing nearly as fast as the adjusted operating income which will cover that debt by a large amount as we talk about reaching adjusted operating income profitability in 2028.
FABER: Right. Does your inability or your unwillingness to issue unsecured debt, certainly at the prices you'd have to right now, in any way, mitigate certain expansion plans you had? For example, building your own data centers.
INTRATOR: Yes. So, look, CoreWeave has so many levers to manage our cost of capital. It has so many levers, including accelerating, slowing down, but we are a success-based business. We are only going to place deals into the market if they make economic sense, including the cost of the components, including the cost of capital, including the cost we, the price we're able to sell at. And so we are targeting margins here, and the margins are expanding, and that is the key to this. The margins for the infrastructure, for the product, for the cloud that we are delivering is expanding.
CARL QUINTANILLA: What does it mean for more equity issuance?
INTRATOR: So, once again, we really do look at all of the tools available to us to ensure that we are raising money in the most cost-effective way possible. We have issued equity in the past through pipes, through other vehicles. We have also issued convertible notes. But right now, the least expensive capital available to us was the prepayments. And you saw us move very aggressively to raise $9 billion of prepayments and deposits because that is the right way for CoreWeave to raise the capital that it needs now. And because our product, because our cloud, because the infrastructure we deliver is so in demand, we were able to turn those levers very, very quickly in order to raise the capital that we need to continue to execute on our business plan. It's fantastic.
CRAMER: Alright. So, Michael, you are an Nvidia shop. You're known as an Nvidia shop. But the day I met you, you were incredibly close to Jensen Huang. First, what has he told you since all this has occurred? Second, is there someone to deal with at this 40,000 person company, 42,000, besides the man who founded it and we all think is the only person that you'll ever hear or see from this company?
INTRATOR: Okay. So, look, Nvidia's an extraordinary company, right? It is not one person. It is an extraordinary company. There are thousands of people over there that interface with thousands of people at my company every day. They integrate on the technology. They integrate on the physical infrastructure. They integrate on the networking. We have a close relationship with them up and down the companies, and they are a, a great partner. They have been a great partner for us for years. They will continue to be a great partner for us. With regards to Jensen, I just don't think it's really appropriate for me, or for that matter, for anyone to really have too much of an opinion here. They've begun a process. Let's let the process run.
CRAMER: Alright. Well, how about the chips themselves? A lot of people felt that these things, wait for five years, would be worthless, including people who told me out of my mind that I thought his chips, this company's chips would be worthwhile. The value of a five-year-old H100 chip, is it worth more than you paid for it five years ago?
INTRATOR: So, this is one of the most exciting aspects of the earnings report that we just went through. It's where I got to talk about the H100s that we installed four years ago will be coming off contract in 2027, and how that technology has now been contracted at a 20 to 50 percent premium over where it was originally contracted. The demand for the infrastructure is enormous and overwhelming. The life cycle of this has been a, this has been an ongoing discussion in the market, and I believe that this should resoundingly put this to bed. This infrastructure will last beyond the depreciable life. And when it moves—
CRAMER: Do you see a two-way market at the compute because of this?
INTRATOR: I think it's one of the tools that the world is building.
CRAMER: Are you helping to do it because you know the country needs it?
INTRATOR: We absolutely are participating in it. We have a lot of things we got to do, and two-way market is definitely part of what has to happen in order for there to be transparency on pricing.
FABER: Yes. I mean, speaking about pricing, I think you raised prices 25 percent in July, another 10 percent later in the quarter. Any pushback at all from customers?
INTRATOR: The customer base realizes that there is a supply-demand imbalance, and they are looking at it from the perspective of what can they generate at their companies from the compute. So, they are happy to pay an increased level because the margins that they are making, and you saw this in the Anthropic paperwork, the margins that they're making, they support higher prices for compute, and therefore we have tremendous demand for this infrastructure. And with that, we were able to raise prices. I think you've seen that universally across the space.
FABER: Yes.
QUINTANILLA: Did—
FABER: Sorry.
QUINTANILLA: Did Hinojosa respond to your letter, and can you address how much of planned capacity might migrate into '28? Because you did say you're waiting for '28 to season, I think is how you put it.
INTRATOR: So, I want to be very clear. We're not, our build plans are not being delayed or affected or, in any way, transformed. The build plan for 2028 is in progress. So, when we say that we're waiting for it to season, we just think that it is better for CoreWeave to wait with the infrastructure and be closer to delivery before we bring it to market to sell it. And that's really how we've approached it. We're trying to compress that timeline so that we're able to get the best prices, the best margins, and the best customers into that infrastructure.
FABER: Carl did reference a letter I believe he sent to the governor-elect in Texas, who's obviously ran on potentially pausing data centers. Any response there and/or any expectations? I think you said the post-election ship was seismic, but it would not have an impact on your business?
INTRATOR: Yes. So, look I want to be clear. That letter got a lot of mileage in the media. But we have relationships, contacts with all of the governors that are in states where we have data centers. And we have in excess of 53 data centers around the world. You know, the letter was picked up in the press aggressively because of the perspective that things are really changing in Texas. When I talk about our infrastructure and our build and our timelines, I'm talking about our backlog, which is an enormous amount of cash that will ultimately cover all of our debt, will cover all of our operating expenses, and drive enormous profit to the company. And that's what the important part is, and we want to make sure that the investing public and our clients know that there is not a disruption in our ability to deliver that infrastructure. And CoreWeave has done an incredible job delivering the infrastructure.
FABER: Do you believe in the overall thesis, though, that if in fact fewer data centers are built it actually makes the compute even more valuable than currently exists?
INTRATOR: So, there is a great argument to be made that if there is, you know, our operating assumption is not a slowdown, but in the case that there is a slowdown, it does make the installed capacity that's operational, the GPUs that are coming out of contract, that, by the way, are fully paid off, right? Our DDTLs have paid off in excess of $4 billion already, and that is wonderful to see that machine work, right?
CRAMER: Okay. So, then, Mike, let me ask you. You're a trader. You've traded. I know you're a good one. Is the best investment that your company could make be the nine and three quarters of October 1st, 2031, which you could buy right now at 81.7? Should you be doing that instead of all this other stuff?
INTRATOR: So, I would be lying if I said that I haven't spent a lot of time thinking about buying back our bonds, right?
CRAMER: Well, then don't lie. Why don't you do it?
INTRATOR: Because I can deploy the capital internal to our company at a higher rate of return. I can get new—
CRAMER: Better than that rate of return? You better get a 16 percent return. I don't see a 16 percent return.
INTRATOR: You saw us guide to an adjusted operating income of between 16 and 17 percent next quarter, and you can look at the trajectory of our adjusted operating returns and realize that it's going to be far in excess of 16 percent. I would love to buy those bonds, but the truth of the matter is we have better uses for capital in our company than buying back the bonds.
CRAMER: Well, maybe those who don't have better uses should be thinking about buying those bonds.
FABER: Yes, perhaps. Can I, real quick, the bears on your stock will point to all the selling from you and your co-founders, saying $3.3 billion since the IPO, $2.4 billion this year. How do you respond to, you know, what are significant sales, obviously?
INTRATOR: So, this is the, this is going to be, first, time I've said this. I've canceled my 10b5.
CRAMER: Yes, thank you.
INTRATOR: It is done, okay?
CRAMER: Thank you.
INTRATOR: I am, you know, any kind of selling from us will just be RSUs, kind of normal tax stuff. But there will be, you know, until there is a material change in our equity pricing, my 10b5 has been silent.
CRAMER: ATM still?
INTRATOR: ATM is different. That's a corporate level.
CRAMER: I know. But why do it though?
FABER: Yes, but that's not him selling.
INTRATOR: Yes.
CRAMER: No, but why not cancel it? If the stock's so cheap, why not let it get, well, get higher?
INTRATOR: The ATM sits on the shelf. We will use it opportunistically if we need to raise capital. That's the right thing for a company to do.
FABER: All right. Thanks for sharing a little news with us too, Michael.
CRAMER: Yes, thank you.
FABER: I appreciate that.
CRAMER: Michael, thank you very much. CEO of CoreWeave, who comes on when the debt's high, low, it doesn't matter, because now it's got prepayments.
FABER: Rain or shine. Rain or shine.
CRAMER: Thank you, Michael Intrator of CoreWeave.
FABER: Thank you.
INTRATOR: It was a great quarter for us. Thanks, guys.
FABER: Sure thing.
For more information contact:
Stephanie Hirlemann
CNBC
