Closing Bell - Oracle Blowout Quarter: Cloud Giant Powers Higher; Plus, CEOs of Chime, CoreWeave and Rubrik talk AI 9/9/25

Episode Date: September 9, 2025

Oracle delivers a stunning blowout quarter driving tech stocks higher. We bring you instant analysis. Apple's iPhone event takeaways with Patrick Moorhead from Moor Insights and DA Davidson's Gil Luria. Exclusive interviews dominate the lineup: Chime CEO Chris Britt discusses fintech growth, Goldman Sachs Global Institute Co-Head George Lee shares insights from Communicopia, and CoreWeave CEO Mike Intrator talks AI demand. Rubrik CEO Bipul Sinha discusses earnings results and a key recent acquisition. Interactive Brokers Chief Strategist Steve Sosnick examines retail trading trends.

Well, that bell marks the end of regulation. Clear Channel outdoor, ringing the closing bell for New York Stock Exchange, Webb's investment doing the honors at the NASDAQ, and stocks modestly higher today, with the Dow leading gains as United Health rallied, all three major averages closing at record levels. Energy and communication services, the S&P sector leaders, materials, and industrials, the laggards. Big Tech, mostly higher with Alphabet and meta, the Mag 7 standouts, Alphabet on pace for its best five-day stretch since 2022. And a green day for the commodity complex, oil, net gas, closing higher, gold losing a bit of

its lustre into the afternoon, but managing to hit an all-time high in earlier trading. And Treasury yields rising that conversely halts a four-day bond rally, the two-year yield bouncing off the lowest level since 2022. That is the scorecard on Wall Street of winner to stay late. Welcome to closing bell overtime. I'm John Ford at CNBC headquarters. And I'm Morgan Brennan in San Francisco at the Goldman Sachs, on technology conference. We have got a huge lineup of guests coming up, starting with Chris

Britt, the CEO of CHIME. We're also going to talk to the CEO of Corwee. That's another recent IPO. Stock hire. Today is the company announces a VC arm targeting AI startups. And sticking with the AI theme, we've got earnings from Oracle due out in just a few minutes. We'll also be getting results from Rubrik and talk to that company CEO Bipple Sinha before he gets on the call with analysts. But let's jump straight in here with the market. as the NASDAQ closes at a record high once again. Christina Parts Nevelis is there at the NASDAQ for us. Christina.

John, we really got a theme today, and it's about tech partnerships, and it starts with Apple's biggest launch of the year, and it's reshaping really supplier relationships across the board. Apple's new AirPods with live translation features initially hammered dual lingo. You could see midday that plunged just post-2 p.m. Eastern, but did recover closing almost 1% higher. Meanwhile, Apple's shift to making its own Wi-Fi and Bluetooth chips cuts out Broadcom, sending shares down over 2% for Broadcom.

And Taiwan Semi does remain Apple's critical manufacturing partner, gaining over 1.5% as the usual winner from any type of iPhone refresh or iPhone event. Elsewhere in partnerships, Super Micro, jumping over 6% on its Nokia AI data center deal combining switching hardware with network automation. I should say, it's closed 7% higher. And then you've got Nebius. Nebius exploded today, over 49% higher after landing a massive 19 billion multi-year deal

to provide Microsoft with data center capacity for AI workloads. These partnerships are really just creating clear winners and losers with AI infrastructure deals driving the biggest moves today, John. All right, Christina, thank you. And now let's turn to Apple, the company unveiling a new lineup of products,

including four new models in the iPhone 17 The iPhone Air included the Apple Watch Series 11 and AirPods Pro 3. Apple essentially shifting the focus back to its moneymaker hardware as it struggles a bit in the AI space. This event hasn't been enough to help the stock today, which closed lower. Joining us now, Patrick Morehead from more insights and strategy and Gil Loria from D.A. Davidson. Guys, welcome. Pat, I look at this line extension as really interesting because it's sort of like adding a new car to the lineup. You got the 17, which is the sedan, and you had the SUV, which is a pro, bigger than some people want maybe, the truck and the pro max. Now you've got a

sports car in the air, less featured, less roomy, but charging a premium. You think it'll work? It's a crapshoot, John. I mean, nothing that Apple brought out here was definitive on anything that can move revenue, decrease cost, take share. And I think the market saw this, and it's at the same place that I am. And if I look at those six things, ASP lifts, accelerating replacement cycle, share shift, new category, product or service, new service attached, lower cost. A lot of thesis, a lot of nice products, John. But if you're adding a, this is not a new category.

I think the first thing that the iPhone Air is going to cannibalize is the iPhone Pro Max versus some premium Android device. Interesting. Gil, do you agree with that? I mean, if you wanted the Pro Max, bigger phone, more horsepower, a lot of camera power. It seems like the air might not be in your sites. But then at the same time, I wonder, is this AI delay from Apple really going to hurt them from here? do you think they did enough with the core capabilities to keep that core customer? Most of the improvements were incremental to Patrick's point. On the air, though, I would say they

are appealing to a different segment. I think the analogy to sports car is a good one. It's leaker. It's sexier. I'm going to guess consumers are going to buy it without a cover because then it stands out. We haven't had an iPhone with a really different form factor in a while. They've coated the back, they coded the front. They're saying you could drop it without a breaking, without a cover. I think that's a new segment. What I would point out in terms of AI is that the live translation feature on the AirPods is AI. If it actually works, and we have to be a little skeptical because it would be very cool,

then it is an AI breakthrough. Let's not forget, Google and Meta did these displays with glasses, and they were a little shaky, at their most recent events. If we could actually use the AirPods for live translations, that's a feature that would actually get people to upgrade. Yeah, I guess that explains the duolingo wobble that Christina Parks and that was mentioned earlier. I also want to mention Oracle results have crossed.

We are going through them. The stock initially dipped. Now it's higher by about 4%. It would be interesting to see what the numbers say there. Patrick, give me your take on Oracle ahead of these numbers and the degree to which they've benefited as sort of a fourth hyperscaler from some of the lack of capacity at AWS, Azure, et cetera. Yeah, that's right, John.

Then being a late entry, I think, surprised everybody, right? We thought the hyperscaler is going to run away with everything. And then Oracle and Larry doubled down on CAPEX and made the investments here. And listen, if Nebius and Corweave can find. customers, I surely would expect that Oracle would find even better customers because they offer more. Their pricing compared to some of the larger hyperscalers is unique in that they have a lower price out the gate and then they upsell you on features where at AWS it's actually opposite. So in a market that was 10 years old, they entered late. They came up, yes.

Hang tight. This stock is taken off. It's up more than 10 percent now. Let's get these results from McKenzie Sagalos. Hey there, John. So it's actually a miss on both EPS and revenue, a $1.47 compared to the street's estimate of $1.48 for EPS revenue coming in at $14.93 billion versus $15.04 billion expected. Now, cloud infrastructure revenue that rose 55 percent year over year to $3.3 billion. This continues to be Oracle's fastest growing business. RPO growth. We are talking about backlog here. This is the cleanest proxy for future revenue growth. That is up 359%. You've got Larry Ellison, the chairman, and CTO, the company saying in the shareholder letter that multi-cloud database revenue from Amazon, Google, and Microsoft grew at a rate of 1,529 in their fiscal first quarter. I'm going to be digging in for operating margin. We got the call starting at the top of the hours. I'll come back to you with

that, but those shares up more than 11% now, John. Yeah, Mac, thanks. Pat, I'll come back to you in a moment, but Gil Loria, I take it that these remaining performance obligations up 359% and this multi-cloud database revenue. We're just talking about them kind of being the fourth position here with the hypers. Is that what has this stock so much higher, you think? Yeah, everybody's very excited about the massive ramp and OCI. But let's not forget, these are not organic or customers. What Oracle has turned into is a neocloud that's there really for overflow capacity from the big hyperscalers. So Microsoft, Google, Amazon are using Oracle, Corweave, Crusoe, Lambda, and others for their excess demand from their customers. So Oracle, to Patrick's point,

prices lower. They have very low margins on this business. So that's why you got to keep an eye on the margins and on the earnings because that growth is coming in with very little margins while the rest of the business is declining and that had high margins. So that's what I would look out for. Patrick, forgive me. I want to go back to you on Apple here because of the expansion of their use of their internally designed modem, the potential impacts there on gross margin and the impact on their ability to put out this line without significant price increases. When do we know if that's really going to work for them? Yeah. So, John, from a customer acceptance point of view, I think that they have already answered unless there are some issues down the line on the higher end

modems used in the pro and the pro max, it's probably not as high as speed as Qualcomm, and the reliability is likely lower, but the power pull is probably lower too. So we won't know until those phones get out there. I feel like there could be a push scenario here where, okay, you make a couple points of margin by doing some of more of your own networking silicon across broadcom and Qualcomm, but you also have this tariff issue that is out there that nobody is really discussing at this point that very much is a reality through Vietnam and also from China. Yeah. We'll see how they perform on that through the holiday season with these new phones as well. Pat Gill, thank you. Now let's move on to synopsis earnings. That stock is lower by about 4% here in overtime. Julia Borson has the numbers. Julia.

John, synopsis missing on both the top and bottom lines with adjusted earnings of $3.39, falling short of $3.74 anticipated, while revenues of $1.74 billion lower than the $1.7.7. $7 billion estimated. The company guiding to Q4 revenue, though, it's fiscal Q4 revenue in a range ahead of the consensus estimates guiding to 2.23 to 2.26. The consensus estimate is $2.09. I'm sorry, $2.9. I just want to point out here a note about the challenging geopolitical backdrop in the release here. The CEO is saying, while I'm proud of how our team navigated external challenges, our IP business underperformed expectations, saying we are taking action to enhance our competitive advantage and drive resilient long-term growth. Stock down about 4%, John.

All right, Julia Borsten. Thank you. Now, take a look at shares of Bill. It's Bill Holdings. Shares spiking more than 5% here in overtime after the FT report's activist fund, Elliott Management, has built an at least 5% stake in the payments automation company.

The report comes one day after another activist fund. Starboard Value nominated for directors to the board. Starboard has an 8.5. percent stake in bill. Morgan. Well, John, take a look at shares of FinTech Chime as well, because part of the summer's flurry of hot IPOs. But after that first day pop, the stock has been sliding. It's now trading below its IPO price. So what is next for the Neobank? Coming up, Chime CEO and co-founder Chris Britt will join me in an exclusive interview as overtime rolls on from Goldman Sachs

Communopia right here in San Francisco. Don't go anywhere. Welcome back to overtime. Rubric earnings are out. Shares spiked higher. Now they're about flat to slightly higher. Let me give you the numbers here. Revenues came in at $310 million versus $282 million expected. Should I give them all here? Yes? No? Yeah, I'll give them all to you. There's an adjusted loss of $0.3 versus $0.34 expected on the guidance. It's also a BQ3 revenue came in at 319 million to 321 million. The range, so the midpoint 320 versus 302 million expected. And then the loss per share that they're guiding to is negative 18 to negative 16 cents versus negative 26 expected. So that guidance on the top and bottom, both better than expected. We're going to get more on the quarter in an exclusive interview with Rubrik CEO, co-founder, Bipple Sinha, right here on overtime,

for the call in just a moment. Morgan. All right. Well, FinTech Company, Chime, rolling out a new premium card, offering cash back, along with perks like fee-free over draft protection, early access to pay and access to exclusive chime deals and more. The company made a strong debut on June 12th when it went public. The stock has been under a little bit of pressure since then, as you can see on that chart right there. But joining me exclusively is Chime CEO and co-founder Chris Britt from Communicopia. Chris, it's great to be with you today. Thanks for having me.

All right, so let's start right here with this credit card, this new offering. What does it bring to the marketplace? And what does it mean for your customer base? We're really excited about the launch of Chime Card, which is just going to make the existing chime banking services even more rewarding. So it's built upon all the great things that you highlighted, such as helping members avoid fees, get access to short-term liquidity on demand, the opportunity to build credit easily just with their everyday purchases, high-yield savings.

and now the ability to earn 1.5% cash back on all of your everyday transactions. And it actually comes with an optional titanium card, which I think is apropos given the Apple titanium iPhone error that just came out, titanium version. So this would match perfectly with that, I think. Oh, there you go. New products from Apple, new products from China. That's right. Direct deposit is really the backbone of your business.

How does that flywheel continue and continue to grow with an offering like this? Well, I think that's the unique differentiation of CHIM relative to many other companies in this sector. It's this deep relationship we have with our members that are predominantly using us as a primary bank account. To give you a sense for that, our average customer does 55 transactions a month. So with our card or just P2P transactions and other money movement transactions. So we have an incredibly deep relationship with our members. And that allows us to essentially have a platform service. When we launch new products and services, they get adopted at extremely high,

rates, such as our recent MyPay launch that we now have about 30% of our members using our ability to get access to your paycheck on demand at any point in time over the course of their pay cycle. So we expect to have similarly very high engagement with this new product as well, as well as future products that will layer on over time to drive this deep, engaged primary account relationship. You're in a very unique position because you're very, very focused on middle income and lower income consumers here in the U.S. what are you seeing from the activity and from the data at CHIME?

Well, we're obviously monitoring all the data and the unemployment reports and all the changes to that. I think at the end of the day, you know, the economy is still essentially at full employment, and that's what we see in our data. We still see a very robust spending and transaction activity happening from our cards. We see our direct depositors spending incrementally more year over year on a per customer basis. And so, you know, we have a very important. We haven't seen any disruptions, we haven't seen any uptick in unemployment benefits, payments, or the sorts of things that we, you know, sort of uniquely see because of this deep primary account relationship that we enjoy.

Yeah, and you've said in the past that you're not necessarily exposed to credit cycles either. I guess that being said, what are you seeing in terms of where we are in this cycle? Yeah, that's right. I mean, we, because of this primary account relationship, most of the spend that we see from our cards, which is how we generate most of our revenue from interchange when kind of. cards get used for everyday purchases, most of our spending is for non-discretionary spending. So even in more challenging economic times, people continue to use CHIME accounts. We saw it during COVID. They got their Stimmy payments into the CHIM account or unemployment benefits. And then when they get their next job, they sign up for direct deposit again, and they use

it for everyday transactions. So, you know, it's tough to say where we are in the cycle. It seems like there is a little bit more pressure. And maybe there is some inflation that's picking up. But even on the inflation side, modest amounts of inflation. is actually good for the business because we're primarily a payments-driven business. So it's a very resilient relationship

and business model that we have because of that deep relationship. And you just talked about payments. I mean, payments revenue really a focus for you and not necessarily net interest margin the way some of the bigger banks are. That being said, how do rate cuts from the Fed factor in here?

Well, rate cuts don't affect us as much as it would a bank. Again, we are a technology company that partners with banks that holds the deposits of our members. And, you know, what you'll see at big banks when interest rates come down is they'll reduce their relatively low rates that they pay on savings accounts today. So I think that'll probably make our offering even more competitive, again, because we're

primarily a payments business rather than a heavy lending business. So quickly then, AI, big topic here, as to be expected. How are you thinking about it? I mean, we're at an incredible moment in time. There's so much hype and for good reason for, you know, the frontier AI companies, the LLMs, the chip companies, and it's incredibly exciting, this AI fuel that they're providing for industry to innovate. And I think, you know, in the same way that in the industrial age, a lot of attention went to the companies that created the energy, the steam companies, the coal companies, and so forth. The winners were also the companies that harness this energy source, this new energy source, and created new innovations and new industries.

And that's how we're looking at it as chime, not at chime, not at chime, not. just to create more efficient services, which we've already reported on the tremendous efficiency in operating leverage from using AI for customer service and that sort of thing. But we're really excited about the opportunity to use AI to make even better consumer experiences. Imagine an application in your pocket that could give you advice on what to spend, what you should do less spending on, how to improve your credit score, how to think about long-term investing. We envision a world where chime plays a very significant role in that, And we think we're uniquely positioned to do that because of this deep relationship we have and the data advantage from that.

Okay. Chris Britt of chime. Thanks for joining me. Thank you. John? All right. Yeah. And speaking of AI for years, Nvidia has been a market darling enriching many people on its way to a $4 trillion market cap. But in the last month, as the NASDAQ has soared to highs, Nvidia has lagged its Mag 7 peers. Have retail investors started to sour on Nvidia? We're going to ask Steve Sosnik of Interactivity.

brokers next. And I want to show you shares of Oracle. That stock is soaring, up 21 and a half percent here in overtime on the back of earnings. The company reporting first quarter cloud infrastructure revenue up 55 percent remaining performance obligations. That's the backlog, up 359 percent. That's more than 4X. We'll be right back. Welcome back to overtime. Invita is lagging its Mag 7 peers in the last month as the NASDAQ has soared to highs. Our next guest gives us the retail take on NVIDIA and why his clients have been buying the dip. Joining us now, Interactive Brokers, Chief Strategist, Steve Sosnik.

Steve, they might not be buying as much NVIDIA, but they're buying a lot of Oracle right now in overtime. So what's your take on how broad this AI story is, especially for some names that we're getting the sense, whether we're talking about Broadcom or now Oracle, that might have seemed secondary before, but now seem to be moving. Well, John, great to see you.

I think there is a little bit of what I'm going to call the broadening out. You know, I've been sort of asserting all along that Broadcom either belongs in the MAG 7 or maybe we can call it the Great 8. I'm wondering out of this point, if Oracle, if they can put up numbers like this and considering their size and Larry Ellison being the second richest guy in the world, I can't come up with anything for nine right now, but maybe some, Maybe that acronym needs to be expanded. But I think all things cloud, all things AI, are still capturing investors' imaginations,

and the fact that you can, you know, the amount of money that's just been put on top of Oracle because of admittedly a very good forecast is just astounding. It shows you the power of these moves. Yeah, from my back of the NACC and calculations right here in overtime, it would be moving from around a $678 billion market cap if this 20 percent, move holds to more than $800 billion in market cap, sort of approaching that trillion dollar mark, where, of course, Broadcom crossed a while back. It seems to be, even though we haven't talked about them as mega caps in the past, perhaps adding to the idea that these big

stocks can continue to power the market, maybe in part because some stocks that weren't as big are now getting bigger. That's kind of what's going on. You know, remember the indices, as we look at them are very, very top-heavy. You have a handful of stocks make up the majority of the weight in the, certainly in the NASDAQ 100, and to a lesser extent, the S&P 500. And so what happens is as these stocks go, so do the indices. Now, the flip side of that was in April when the markets hit a bit of an air pocket, shall we say, the bigger the stock, actually, the worse it did. You know, the Mag 7 drastically underperform the NASDAQ 100, which underperformed the SPX, which underperformed the equal-weighted S&P 500.

So, you know, live by the sword, die by the sword. But right now, this momentum is still continuing to carry stocks forward. And it proves to us, again, that it's not necessarily what the company says for the quarter. It's what they say going forward because Oracle on a quarterly basis was met at best. But in terms of the guidance, the market's obviously loving it. And in a way, Steve, does this Oracle result help to bolster or justify retail investors buying the dip on Nvidia because Oracle says in its release, multi-cloud database revenue from Amazon, Google, and Microsoft grew at an incredible rate of 1,529% in Q1.

I mean, part of the multi-cloud story, it's linked to the AI story, and one could figure if they've got demand on that, then they've got demand for Nvidia chips as well. Well, it's one giant ecosystem. You know, remember that, that Nvidia's largest customers are four of the other Mag7 stocks. And so if Oracle, you know, if Oracle can basically take that same sort of role as a cloud, as a cloud component, you know, that Nvidia has as a chip provider, obviously that's going to provide a lot of heft for Oracle because, you know, these are the companies that are earning the most money and have the most money to spend. and if they're spending it with Oracle, that's great news for that company shareholders. And I want to mention Oracle typically, technically doesn't give guidance until the call, but with this remaining performance obligations number that they've given,

they're sort of giving a glimpse into the pipeline. We'll see if the stock moves on the guidance as well, either up or down. Steve Sosnik, thank you. Morgan? Yeah, Monster move for Oracle right now. Thanks, John. Coming up from the Communicopopia Conference, we're going to be joined by George Lee, co-head of Goldman Sachs Global Institute, one of the top AI experts on Wall Street. We're going to speak with him about AI's power problem, how to solve it. We may ask him about some of these AI-related earnings from the likes of Oracle as well.

A lot to get to. We'll be right back. Welcome back to overtime. Another record day for the markets. Dow, S&P, NASDAQ, all closing at highs. Oracle, a monster gainer here in the after hours, up 21% missing on overall earnings and revenue. Who cares? Investors are seizing on what's called remaining performance obligations. That's future revenue from signed deals.

That number rising more than fourfold to $455 billion. And Oracle's multi-cloud revenue from three of the biggest customers, Amazon, Google, Microsoft, grew more than 1,500 percent in Q1. Founder and exec chairman Larry Ellison said he expects that overall number to grow substantially every quarter for several years. Now we've got breaking news out of Washington. Our Megan Casella is at the White House. Megan. John, some big news just now from the Supreme Court agreeing to hear the Trump administration's appeal of the case that invalidated the majority of his tariff agenda. Remember, these are the AIPA tariffs that are in question nearly all of those country-specific tariffs we've seen so far this year imposed. that Trump administration lost this case in its first round of appeals just a couple of weeks ago,

appealing it now to the Supreme Court. The Supreme Court saying they do agree to hear this case and they also agree to the administration's motion to expedite this case. That means we are going to see this on quite a fast-tracked timeline here. The Supreme Court laying out in a short brief saying that they are not only agreeing to hear the case, they are looking for opening briefs to be filed by Friday, September 19th. That's next Friday. And then a long timeline here that allows for the cases to be set for argument the first week of the November 2025 argument session. So in the first week of November, John, we will see both the administration, its lawyers, as well as the plaintiffs in this case arguing before the Supreme Court on the, on the

legality of the president's tariff agenda. John? And Megan, there's a big constitutional question here, right, on the extent to which Congress can delegate its tariff powers to the president? That's absolutely right. That is the question at the heart of this case. The Constitution initially delegated tariff powers entirely to Congress. Over the past several decades now, Congress has slowly given some of that more and more of it to the president, usually in specific instances. And the statute that the president used here, AIPA, the International Emergency Economic Powers Act, doesn't even include the word tariff within it.

So the way or the reason, I should say, the administration has lost the first two rounds, the initial case and the first appeal is by saying that this statute doesn't go far enough to allow him to impose tariffs of such sweeping nature as he's done. That's what the Supreme Court will be looking at. The administration, of course, has other options to pursue tariffs if they lose this case, but this one does regard the vast majority of everything we've seen the president imposed so far. Yeah, key question. What constitutes national security here and what that means in terms of the powers for the executive branch? Megan Casella and Washington, thank you. Well, as As tech continues to dominate the broader markets, and as an artificial intelligence continues to lead the narrative, both for investors and for corporate America, there's no one better to talk to about it than George Lee.

Now, he is co-head of the Goldman Sachs Global Institute, where he focuses on all things AI. He's widely considered one of the top experts on AI on Wall Street. He joins me here now, George. It's great to have you. Morgan, great to be with you. I've got to start with Oracle. Last I checked, those shares are up 20% right now. This astounding RPO number, up 359% year.

over year, almost half a trillion dollars now. How does it speak to what we're seeing in AI and, I guess, the shifting landscape in terms of leadership positions? Well, it's certainly an extraordinary market event for a company of that scale, for sure. I think it just reflects this situation we find ourselves in where the infrastructure boom is so profound, and yet we find ourselves in a moment of capacity constraint. There just aren't enough compute cycles to meet the demand for tokens in the world. And so Oracle's booming business, the booming business of all these hyperscalers, They're delivering these computational cycles.

It's just a commentary on what we see in the kind of demand and consumption profile of intelligent tokens. How long and how sustainable is this spillover effect when we talk about capacity constraint for the oracles or the core weaves or some of the others of the world that are taking on some of this capacity that the hypers themselves can't handle right now? Well, that's obviously the trillion-dollar-plus question. There's no doubt.

And, you know, I think it's a consequence of, will this demand profile maintain the current, trajectory, the steep trajectory that we're on. How quickly can we deliver these cycles and what other constraints may come to bear? I'm sure we'll talk about energy, which is emerging is one of the principal constraints here. And I think it has a lot to do with whether and when we'll find product market fit for these AI technologies within enterprises and with consumer use cases. We've seen great early progress, but that has to be sustained and accelerated, I would argue, to meet the moment of all of this spent.

So let's talk about it, because you spent a lot of time focusing on it, you've written papers on it, but this intersection of power and AI, where are we at in terms of that cycle, that build-out, what needs to happen to fulfill the energy piece of this more quickly? Yes, I said it's really emerging as a principal constraint for the ability to deliver the infrastructure that produces these intelligent tokens.

And that's in many dimensions. Generation, transmission, supply chain, human labor. actually we're running short of electricians that are capable of delivering this kind of this kind of power and I think there are two two principal things to remember here one is like your father's cloud data center was architected for five nines of availability and very low latency because all of us abandon our shopping cart in point six seconds second important thing to remember is that we build power in this country for peak demand those days where it's 100 degrees in New York or Houston or whatever and thinking about the those two foundational elements, we're arguing that there is a way to brook that temporal gap. I mean, this ultimately, we will produce enough power assets, will upgrade transmission so we can meet this moment. But right now, there's a bit of starvation. Our belief is that the ability to think differently about the availability of these AI data centers is going to be a big difference maker. There's a paper out of Duke that talks about the ability to curtail workloads.

So in those moments when it's 100 degrees in Houston and that power is required, you can manage down the demands on your AI data center, operate at three nines of ability, a slightly lower latency, and therefore take down those workloads and then harvest the electrons that happened during more of the slack period. Now, Duke argued that there's some 75 gigawatts of power that you might be able to harvest from this workload discrimination and demand response. We think this is a really interesting way to brook this divide. But make no mistake, there is going to be a long-dated, persistent investment cycle in power necessary to power this revolution. So basically the data centers themselves become smarter. And we know there's a whole startup ecosystem that's focused on this right now. So I am curious because we had Allie Godsey from Databricks on the show yesterday.

And he talked about the fact that maybe there's been at least initially over-excitement, over-spending, over-investment on the promise of AI, and that that seems to be coming back down to Earth, but simultaneously what he's seeing right now is that there is more adoption, there is more realization, at least the beginning of realization of return on investment. What are you seeing from your key vantage point, especially as many of these companies are starting to come public? Yeah. Well, Ali, obviously, a super informed and very balanced observer, so I take super seriously what

he says. You know, we have found product market fit in the coding space, engineering and coding, and we find that at Goldman Sachs. Our engineers are using these coding capabilities and agents to produce more code. We found product market fit and customer support. Mark Benioff talked at length about that in his most recent earnings call. I would argue we found it in the advertising space. And, you know, I think it's going to be a steady, slow accumulation of use cases

that consume these tokens in the enterprise. There's a great paper out of Princeton that exhorts us all to treat this as a normal technology. While it has its magical qualities, it's a technology where the diffusion will be slow. Enterprises will take time to find product markets. fit. We'll have to change our habits as workers. We'll have to shift our workloads and our workflows. All that's going to happen. It just may take a little bit longer than we would like. All right. George Lee at Goldman Sachs. I appreciate the conversation. It's great to be here with you at Goldman's Communicopia Conference. Thanks for being here.

All right. Great insights. Yeah, Morgan, thank you. In the meantime, rubric shares are volatile right now, shooting higher initially, now giving up the day's 3% gain. Second quarter numbers out earlier this beat handily on the top and bottom lines, as did the guide. Stocks up about 50% in 2025. Joining me now for an exclusive interview before the analyst call is Rubik co-founder and CEO Bipple Sinha. Bipple, I got to draw a bit of a parallel here between you guys and Oracle in that what's happening in AI, what we've seen with some of the biggest tech companies out there, is having ripple effects throughout the enterprise ecosystem.

How is it affecting resilience, recovery, and cybersecurity? Great to be with you again, John. AI is really forcing enterprises to transform their infrastructure, transform their applications, adopt more cloud. And as they are doing these transformation, they need cyber resilience. And that's why cyber resilience is the number one cybersecurity categories, because everybody is worried about successful recovery from cyber attack and keep their applications up and running. Because if your data doesn't have integrity,

or your data doesn't have availability, AI is almost useless. The stock bouncing now off the lows of the overtime session, now down about just 1%. Tell me, you've been balancing your revenue growth and your cash flow as well. How are you thinking about that? How are you able to do it as some of the customers out there in the enterprise are very focused on where they're willing to spend money?

We had an exceptional quarter. If you look at our top line, we grew 51% revenues to almost 310 million. We generated 57 million in free cash flow, 19% margin. So we are really focused on winning this very large market opportunity that is in front of us in terms of cyber resilience. What every enterprise is realizing is that cyber attacks are inevitable. and they need to have cyber resilience to be able to keep their services up and running because digital trust is key to AI and digital trust comes from understanding that in spite of a cyber attack, the businesses will be up and running.

And that's why customers continue to prioritize rubric cyber resilience. But we are also helping our customers accelerate their AI journey. We acquired a company called Predibase in the AI fine-tuning and model-serving platform, and we announced Agent Rewind that will actually undo actions of a misbehaving agent. So we are really focused on delivering cyber resilience and accelerating AI journey for our customers. And what does that acquisition, what do your sales motions do for your share of wallet and for the net revenue that you're getting over time? Every customer has AI proof of concept, and they want to take those proof of concept into production

and value realization, but accuracy, cost, hallucination, and operational management of agents are important. And that's where we are focused on. In fact, what we are seeing is that our customers are really worried about misbehaving agents or having autonomous action that is negative to their business. And that's where the agent rewind that we launched, which is coming out later in the year, is going to help businesses confidently do AI transformation. All right, Rubik's CEO Bipple Sinha.

Thank you. That earnings call kicking off in about 15 minutes at 5 p.m. Eastern, so we'll let you get to it. Morgan. All right. Well, Corweave, just announcing a new venture fund to invest in AI startups. We're going to hear from CEO Michael and Trader about that fund and get his reaction to Microsoft's deal with Nebius. When overtime returns, the stock shot higher today. We're also going to talk to me about this whole idea of neoclouds and overcapacity and what that means for Corwee.

Stick with us. Welcome back to Overtime, CoreWeave launching CoreWeev ventures today to invest in AI startup. Shares closing higher by more than 5% and actually trading higher here in overtime. Joining me right here right now in San Francisco Goldman's Communicopia Conference is Michael and Trader. Corweave co-founder and CEO. And Mike, it's great to speak with you. It's great to be here. Thanks for having.

I want to get into the Ventures Fund. But first, given the fact that last I checked shares of Oracle were up 20%, it speaks to perhaps this capacity. we're seeing more broadly in the marketplace for AI capabilities. How is it affecting you? Yeah, look, you know, the move in Oracle that is being driven by their backlog to deliver compute, other deals that have been announced within the last 24 hours, all of this is really just a reinforcement of the way that we have been building and focusing our company

on being able to provide the infrastructure for a group of companies that are virtually insatiable in terms of their demand for compute. And you're seeing it across the space. It is really an incredible commentary on the macro environment that we're building our company. Yeah, and of course you just mentioned other deals in the last 24 hours. The Microsoft deal with Nebias getting a lot of attention as well. I want to get your thoughts on that, especially since Microsoft is a key partner of yours. Yeah, look, it's a great deal.

You know, it's a very large-scale transaction. once again, a representation of the size and scale of the demand that we're seeing across the space as the most important companies throughout the lab community, the most important companies across the cloud and AI native communities are trying to get the infrastructure that's required to be able to drive the potential of what AI is going to be. And it's just really, like I said, It's an incredible representation of the macro environment that we're operating. What does it take for CoreWeave to get to a place of meeting the capacity, having the capacity to meet the demand?

And I guess just as importantly, when there's so much money and capital being raised by so many different companies right now, is there capital for you to be able to do that? Yeah, so look, we've been incredibly successful in accessing the capital markets. We've been groundbreaking over the last 18 months. we've raised in excess of $25 billion. We have a great team that coordinates accessing the debt markets as we bring in successful transactions to be able to build the infrastructure

to deliver to our clients. They depend upon us to be able to do that. And it's really been one of the engines of success for our company. What the market is really internalizing over the last couple of days is just something that we have been trying to communicate, I guess sometimes I'm good at it, sometimes I'm bad at it, but just the companies that are building artificial intelligence, that are building the foundation models, that are integrating these models into the enterprise business community,

they just cannot get enough compute to be able to serve their clients, to be able to train their models, to allow artificial intelligence to drive their businesses forward. And with that behind us, with that tailwind, we are really well positioned to, take advantage of that across the capital markets, across the infrastructure markets, and with our partners and clients. Yeah. And speaking of partners and key investors, NVIDIA, you've started to already see some of those first GB300 racks from NVIDIA.

Should we expect that to ramp here in the final months of the year? And what does that mean for this capacity? Yeah. Look, you know, we work very, very closely with our suppliers. We work very, very closely with our clients to be able to deliver infrastructure. And, you know, the ecosystem at large really does depend. upon our ability, and others, to deliver infrastructure at a scale that allows the building to continue at a pace and magnitude that the world is moving right now.

You are going to see massive scaling of infrastructure as we move through Q3 and Q4, as we bring on just phenomenally large blocks of GB200s, of GB300s, as we bring on the networking required to make this compute function effectively. as we deliver, you know, storage solutions. It's just, it's really exciting. It is a moment in time where the size and scale of what's going on is just sort of mind-boggling. So how does this new venture arm fit into this, especially since I would imagine it allows you to not only seed the startup ecosystem, but also potentially cultivate future customers and

gives them access to some of this capability? You're really hitting on a lot of it, right? The idea is that for the environment to be healthy and robust and resilient, you need to be able to allow new companies, new entrepreneurs, brilliant technologists of the future to be able to bring their ideas and turn them into companies, turn them into startups, turn them into the technology of the future. And in order to do that, they need to be able to get access to compute. and by partnering with them, by working with them, by investing in capital, by investing in

access to compute, doing all those things we can wrap our arms around them and help them and help steward them through what is a tenuous component of the entrepreneurial journey. I know this well. And so, you know, we're really well positioned to pick these companies up early and to help them as they kind of begin to establish themselves as important components of the future of what AI will deliver, whether it's on the application layer or it's, you know, anywhere else within the ecosystem. Okay.

And, of course, markets like it. Stock finished up 7% today, up another almost 5% as you're talking right now. Mike and Trader, of course, great to speak with you. Thank you so much. I really appreciate the opportunity. Well, that does it for me here today, live, but don't miss. My exclusive interview is Service Now CEO Bill McDermott. That's tomorrow 2 p.m. Eastern on Power Lunch.

And right here, same time, same place. tomorrow on overtime. I'm sitting down exclusively with Goldman Sachs chairman and CEO David Solomon. John, it'll be a wide-ranging discussion. You do not want to miss that. Two big names. Morgan, looking forward to that. Meantime, here in overtime, still, we've got this big move in Oracle, up nearly 23 percent right now on results. The headline numbers, unremarkable, but the company reporting blowout numbers for its revenue pipeline and its cloud unit. Backlog is up 359%. That's more than 4X. Cloud infrastructure up 55%. Its database lets customers operate across multiple clouds up an incredible 1,529%. With me here now is Ben Reitzis of Mellius Research and CNBC Senior Markets

Commentator Mike Santoli. Mike, you might argue that they should have pre-announced this number, but I guess the top and bottom line was unremarkable. They had this in their pockets. It was all in the future quarters. On one level, first of all, is the latest AI scare over? I mean, you know, We have had a lot of these little kind of gut checks and rethinkings along the way. It's remarkable that the market, at least right now, is willing to add $150 billion to the market value of Oracle on this news. But then you consider, that's a 4% or 5% move in Microsoft. It's almost just putting Oracle in the next league in terms of the size of the market it's addressing in an accelerated way. Don't know if this is going to hold.

We've seen these massive after-hours pops on some of these results, you know, not necessarily. persist. Microsoft actually gave its back after earnings, but I think it's, it still shows you that the street is still buying into the long-term story for Oracle. Ben writes, this a quarter ago, you were rightly focused on remaining performance obligations, that pipeline, and how good it looked for Oracle. There's a distinction here between cloud and multi-cloud and AI, but this chatter about are the AI applications paying off yet? That's different from whether customers are investing in that core cloud transformation and that data transformation, no? Well, John, I mean, I could answer that, but, you know,

nobody'd be spending like this with a backlog like this if AI wasn't real. So, you know, I know it makes good media to kind of say maybe AI isn't catching on and some folks are saying it's a bubble. This is a very historic kind of print right here from Oracle with this backlog. The street was looking for about $180 billion in RPO, and they're talking about a number that is a multiple of that. That is astounding, and I think that this is one of those nights, like, you know, you remember where you were in the bubble, you know, in the 90s where something happened, and this is one of those nights where, you know, the bubble may be starting, you know, in a year or so. But we just had one of those eye-popping weeks with Broadcom and this. I'm literally bolting myself to my chair from this RPO number.

I literally have to. This is a historic night. This is really good news. You mentioned Broadcom. How much of this is unique to Oracle and how much of this is a signal about overall AI demand and the broader market? This is everything. So it is a big signal about AI demand. This is a big value.

It could be a big, like, the big question on the call is going to be like, how much this is open AI and Stargate and is Stargate real? How do you know? How do you know, like, you know, how it's going to ramp and all that? But they did give revenue guidance all the way through FY29, which is stunningly above the street, you know, like 50% above the street for each year, basically, all the way through 29 for the cloud unit only, which is really stunning. So they have the visibility, John,

and it looks like they think AI is real, and there must be a lot from Open AI and a lot going to Broadcom and a lot going to Nvidia here. All right, Ben Wrights. Ben Wrights is from this religious research. Thanks for getting on the phone with this big Oracle move.

Mike, how much does this roll over, perhaps, into the broader market tomorrow? Relatively measured moves in the immediate kind of spillover names we were just showing there, so a percent and a half on Nvidia. Yeah, I do think it just reinforces this idea. You've got to be really careful about calling the peak.

Even if AI is being over-invested right now, it's going to last for a while. We're not going to know if these guys are right or wrong for some time. Yeah, so much for that AI is overdone. That's going to do it for overtime, though.