Nvidia Puts Another $2B Into CoreWeave, Offers New Chips

26 Jan 2026 · 44 min · 17 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

Podcast Summary: Bloomberg Tech - Nvidia Puts Another $2B Into CoreWeave, Offers New Chips

Episode Overview

  • Hosts: Caroline Hyde (New York) and Ed Ludlow (San Francisco)
  • Main Topics:
  • Nvidia's $2 billion investment in CoreWeave
  • IonQ's acquisition of SkyWater Technology
  • Upcoming earnings reports from major tech companies like Meta, Apple, and Tesla

Key Discussions

Nvidia's Investment in CoreWeave

  • Investment Details:
  • Nvidia is investing an additional $2 billion in CoreWeave to support the build-out of AI factories.
  • The deal includes purchasing CoreWeave's common stock at $87.20 per share.
  • CEO Insights:
  • Jensen Huang, CEO of Nvidia, emphasized that this investment is a small portion of CoreWeave's overall capital needs and is not indicative of circular financing.
  • Huang stated that CoreWeave is a strategic partner that aligns with Nvidia's technology vision.

IonQ Acquisition of SkyWater Technology

  • Acquisition Overview:
  • IonQ, a quantum computing company, has acquired chipmaker SkyWater Technology for $1.8 billion in a cash-and-stock deal.
  • CEO Niccolo Di Masi's Insights:
  • This acquisition will accelerate IonQ’s R&D roadmap and enhance their ability to supply the quantum technology industry.
  • Di Masi views this as a strategic move to strengthen relationships with the U.S. government and enhance domestic semiconductor manufacturing.

Upcoming Earnings Reports

  • Market Implications:
  • Major tech companies, including Meta, Microsoft, Tesla, and Apple, are set to report earnings this week.
  • Ed Ludlow noted optimism in the markets ahead of these earnings, particularly regarding AI valuations.

Market Sentiment and Reactions

  • General Market Trends:
  • The Nasdaq 100 is experiencing its longest winning streak since December, amid growing optimism in the tech sector.
  • Discussion among analysts about the sustainability of current market valuations, particularly in AI.

Microsoft’s AI Chip Developments

  • Second Generation AI Chip:
  • Microsoft is rolling out its second-generation AI chip, which aims to reduce reliance on Nvidia's hardware.
  • Initial units are operational in specific data centers, with future rollouts anticipated.

Tech Industry Challenges

  • Investor Concerns:
  • Concerns about circular financing, with some experts suggesting that Nvidia’s investment in partners like CoreWeave may raise questions about market independence.
  • Potential overbuilding in the compute market as companies ramp up investments.

Key Takeaways

  • Nvidia's Strategic Moves:
  • Nvidia's investments signal continued confidence in AI technology and partnerships that can drive innovation in the sector.
  • IonQ's Growth Strategy:
  • IonQ’s acquisition indicates a proactive approach to enhancing capabilities and establishing a market leadership position in quantum computing.
  • Market Dynamics:
  • The upcoming earnings reports will be critical for assessing the health of the tech industry and gauging future investments.
  • AI and Quantum Future:
  • Both Nvidia and IonQ are positioning themselves at the forefront of AI and quantum computing, marking a significant technological shift.

Conclusion The episode of Bloomberg Tech provides insights into significant financial moves within the tech industry, emphasizing the importance of strategic investments and the upcoming earnings that could influence market trends. The discussions on Nvidia and IonQ highlight how technology firms are navigating the complexities of innovation and market expectations.

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

NVIDIA's $2 Billion Investment in CoreWeave

0:45 to 2:36

Discussion about NVIDIA's significant investment in CoreWeave and its implications for AI.

“And the markets are up ahead of those all-important earnings tests.”

The Role of AI Chips in Microsoft

2:36 to 4:34

Insight into Microsoft's strategy regarding AI chips and its competition with NVIDIA.

“Why dig even deeper into their pockets to give money to Corweave at this time?”

Market Reactions and Earnings Preview

4:34 to 11:20

Analysis of market reactions to tech investments and an earnings preview for major companies.

“But the reality is, like all the other hyperscalers, Microsoft still depends on NVIDIA.”

IonQ's Acquisition of Skywater Technology

11:20 to 14:00

Discussion on IonQ's acquisition of a chip foundry and its impact on quantum computing.

“I don't know how true that is, but I think we're going to get a glimpse into how popular or whether or not businesses are paying up for enterprise licenses, and that's always a concern.”

Skywater's Role in U.S. Quantum Growth

14:00 to 19:19

Discover how Skywater is positioning itself in the U.S. government's quantum initiatives.

“Does that help build and leverage the relationships you already have with the U.S.”

EU Probes XAI and Grok's Risks

19:20 to 22:21

Learn about the EU's investigation into XAI's chatbot risks and its implications.

“The reason why this is so exciting for us is that with Skywater, we can make quantum computing a mass market reality.”

Defense Stocks Surge Amid Geopolitical Tensions

22:22 to 26:30

Explore the factors driving the rise in defense stock prices due to global instability.

“In fact, they've soared as governments ramp up spending on heightened geopolitical tensions.”

Power Grid Strain from Winter Weather

26:31 to 28:00

Understand the impact of winter storms on the U.S. power grid and rising energy prices.

“Power grids, they're expected to endure increased stress from really brutal cold in the wake of the damaging winter storms.”

Energy Authority and Power Supply Measures

28:00 to 29:05

Learn about the Energy Department's recent measures to ensure power availability.

“And so they're basically trying to take out all stops.”

Base 10's Role in AI Models and Infrastructure

29:39 to 31:11

Discover how Base 10 supports companies in running AI models at scale.

“AI inference startup Base 10 Labs has raised$300 million.”
Show all 17 chapters

NVIDIA's Partnership and Market Position

31:11 to 33:09

Explore NVIDIA's investment strategy and its role in the AI market.

“Well, I'd say this is a very secular bet for us.”

Customer Strategy and Pay-As-You-Go Model

33:09 to 34:55

Understand Base 10's customer alignment and pricing strategy.

“Well, I think so far it looks like NVIDIA is a pretty strong winner at the chip level.”

Attracting Talent and Market Challenges

34:55 to 37:01

Learn about the challenges of attracting talent in the AI inference market.

“And so that's actually a very important part of our business.”

Base 10's Ecosystem and Support for Startups

37:01 to 37:51

Discover how Base 10 helps startups through their incubator program.

“But, like, is there an opportunity here for you to look at the rest of your portfolio and your historic investments and say to some of these companies, like, you guys should work together?”

Tech Headlines and AI Investment Landscape

37:51 to 40:53

Listen to the latest tech headlines and insights on AI investments.

“Caro, there are so many more tech headlines out there today.”

NVIDIA's Performance and Future Predictions

40:53 to 42:00

Analyze NVIDIA's market performance and the outlook for AI profitability.

“And it looked like maybe the startup or maybe a practice run for the AI bubble popping.”

Analyzing AI Trade and Capital Expenditure

42:00 to 43:00

Learn about the significance of capital expenditure in the AI sector and its impact on NVIDIA.

“the overall sort of expansion of the AI trade is giving bulls sort of more room to run.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:11Bloomberg Tech is live from coast to coast with Caroline Hyde in New York and Ed Ludlow in San Francisco. This is Bloomberg Tech. Coming up, NVIDIA invests an additional$2 billion in CoreWeave, expanding their partnership to accelerate the build-out of AI factories. Plus, IonQ continues its acquisition spree as the quantum company buys chipmaker Skywater technology in a cash-a-stock deal worth$1.8 billion. And gearing up for big tech earnings, Meta, Microsoft, Tesla and Apple all set to report earnings throughout the week. We have the preview. And the markets are up ahead of those all-important earnings tests.

0:53We want to see whether there's vindication, whether revenue and the profit lives up to some mighty valuations, particularly in the AI space. But we're up for a full straight day on the Nasdaq 100, longest winning streak since back in December when we had a five-day run heading into the Christmas period. So there is optimism, but not if you look at the US dollar, not if you're seeing the search for safety in gold. But we stick to our knitting here in tech. And what have you got, Ed? We've got a deal and it's for NVIDIA to invest an additional$2 billion into CoreWeave. They're going to buy CoreWeave's common stock at$87.20 a share.

1:27There is a technology sharing part of this, but the bigger picture is to accelerate the build out of capacity. Bloomberg last night speaking on the phone with the CEOs of both companies. Bloomberg's Ian King, who conducted that conversation, joins us now on set. I'm going to get to the circular financing bit first. We asked him, is this circular financing, him being Jensen Wang, the CEO of NVIDIA, his answer was pretty simple, that he's saying this is a very small proportion of what CoreWeave's going to need to raise on an ongoing basis to build out capacity. And it's in line with what they invest in lots of other companies.

2:03Take it from there. Yeah, I mean, he said, look, whether it's CoreWeave, whether it's XAI, whether it's any of the others that he's, OpenAI, obviously, that he's agreed to give a lot of money to, he's like, this is just a small portion. This is me just helping them out a little bit and showing faith in what they're trying to do. They're going to have to go and raise this money themselves. The market will basically sort this out. This isn't circular financing, was his argument. And it felt like the Corweave CEO, Michael Entrator, came back and said, look, it's a drop in the ocean compared to the amount of infrastructure spend that we're having to do.

2:35But where do they strategically align? Why dig even deeper into their pockets to give money to Corweave at this time? Yeah, I mean, Jensen said that this is really a technology partnership, that CoreWeave is a very early adopter of NVIDIA systems, tends to do them in the way that NVIDIA thinks they should be done. And this kind of serves as like a reference platform, so they like that. And also, he just thinks there's a lot of demand out there, so the faster we can go, the better things get. There's also CPU. Put the GPUs to one side. What Jensen Huang really wanted to emphasize, as far as I can tell from reading the transcripts of your conversation, is that Vera CPU is now going to be a standalone infrastructure product.

3:19CallWeave's relevant here. Just explain what he told you on the phone. Yeah, I mean, as you know, Jensen never says anything without having carefully thought it through. So he really wanted to deliver a message that this is a standalone product as well as being part of the infrastructure. You remember previous versions have been sort of integrated into NVIDIA's computers. Right, one of six chips. Right. Now, hey, you can use it on your own. if you want, and that makes it a direct rival to some big products out there. And real quick, Caro, AMD under pressure, Intel also under pressure. Ian, is that a surprise?

3:49It shouldn't be, no. Right. Ian King, we thank you so much for breaking down a really important story for us. And look, there's another one in the world of AI semiconductors. Microsoft is rolling out the second generation of its own AI chip. So C shares are up about a percentage point, not moving much on the news that is just dropping. and Bloomberg's Matt Day is a person who covers Microsoft for us, breaks that news. How important is it that yet another company is building their own AI chips? Well, for Microsoft, it's really important. They got a bit later jump than their biggest rivals, Amazon and Google, in building their own.

4:21Silicon, they see this as an important way to reduce costs, to find another source of availability. So this is going to be a really big test for their chips unit and for their cloud business. We write in our Bloomberg story that it could eventually provide an alternative, right, to NVIDIA's hardware. But the reality is, like all the other hyperscalers, Microsoft still depends on NVIDIA. Just explain its scale of its compute and what it does rely on NVIDIA for. So NVIDIA is the workhorse that powers the vast, vast majority of the AI workloads in Microsoft's data centers. If you're using OpenAI as a chat GPT consumer, chances are you're pinging an NVIDIA chip running in some Microsoft data center somewhere.

5:01Now, like everybody in the industry, Microsoft would love to have some other options. hence this effort with Maya to try to build a better mousetrap, one more suited for their data centers and one that ideally could reduce some of their costs in the long term. Tell us how it's rolling out, where it's rolling out first, and ultimately when customers are going to be using it. So it's not clear when customers are going to get their hands on this. The first units are live today, Microsoft tells us, in data centers near Des Moines, Iowa. More are coming to the Phoenix area, but we don't have a date yet for what a wider rollout looks like or, you know, sort of crucially, what availability looks like.

5:34Is this going to be a global rollout? Is this more of a kind of R &D step and future chips are going to are going to sort of carry the mail for them? So it's kind of yet to be seen how broad they take this one. Bloomberg's Matt Day out of Seattle with the latest on Microsoft's latest AI chip efforts. Let's talk through the market reaction to all of this with Kim Forrest, CIO and founder of Boca Capital Partners. And go back to the NVIDIA core weave deal. You know, it's a driver in markets this morning. the explanation from NVIDIA CEO Jensen Wang is that it is not circular financing because it is a small portion of CoreWeave's overall capital requirements.

6:09Are you concerned about that? And what do you make of his response? Sure. Well, of course, it's circular financing. You know, CoreWeave, as well as everybody else, even Microsoft, is a big customer of NVIDIA's. So if NVIDIA is investing in its customers, by definition, it's circular financing. It is a good point that, you know, two billion is probably a rounding error for NVIDIA at this point. But it's also true that CoreWeave has to prove itself to the market and raise its own money. Kim, we're saying circular financing, the term, as a bad word, a bad phrase. Is it? Right. Yeah, kind of. You know, Jensen points out that CoreWeave is going to be one of their reference accounts.

7:00What you ideally want in a reference account is somebody that has worked with you, is kind of pure, and has no inside kind of relationship with what they're recommending other than they use the product, they love the product. And, you know, this kind of sullies the relationship that CoreWeave has so much investment from the company that they're giving a reference for. Meanwhile, many would say it's just because we cannot build compute fast enough. And in many ways, NVIDIA, more than anyone, sees the opportunity. And it's just piling money behind this to be able to get the end demand that it needs for the products that it's currently building.

7:41Kim, are we likely to see through these sorts of technological partnerships as well as financial partnerships actually get the compute out at this high speed that we need? Sure. Well, you know, we're going to see them until we don't. I'm not trying to rain on anybody's parade. But at this point, it's all kind of, you know, up in the air. Right now, there is a huge demand for compute. But like everything else in the world, somebody is going to overbuild and And it's going to be jarring and disappointing. I'm not saying that that's any time in the future. I'm just pointing out the obvious. That, you know, yes, this is a very symbiotic world with customers and providers and competitors all kind of working in the same pool.

8:27It's messy and it's like real life because it is. But, you know, just be warned that this is not, you know, trees don't grow, you know, well, trees grow into the sky. but not forever. We'll put it that way. Well, they're not going into the sky today for Intel and AMD because both of them under pressure that maybe here's Jensen coming for their piece of the pie, the CPU part of the pie, that everyone was very excited for them. Indeed, Intel of late in the run up in their stock. Does Vera brand the CPU on its own prove that it's a key competitive to them too? Well, I think at first it's going to work in its own environment and it may become a competitor, sure.

9:08I mean, I think the takeaway from all these announcements over the last, I don't know, 48 hours is that everybody's innovating. And that is excellent. I love competition. I'm a capitalist. Innovate. Try to outcompete. Do it. Live that way. But the fact of the matter is the CPUs that are the workhorses in just plain old cloud computing aren't necessarily going to be interesting or aren't necessarily going to look to NVIDIA. NVIDIA loves the highest priced product. That's what it does. And I'm sorry, the workhorse people need reliability. They need it to look the same across their platform. They need a lot of stuff that a new CPU chip isn't necessarily going to be interesting to them.

10:00Now, in the world of AI data centers, sure. But in the old-fashioned ones that are going to have to persevere for decades, I don't know that that's really a competitor, NVIDIA. That's an interesting observation because the GPUs in the AI context are ARM-based. And so if you're running those GPUs of x86, you're going to have to do some special software work to make them match. Let's put that to one side. It is a massive week in the world of technology from earnings alone. We also have a Fed meeting on Wednesday. Of those names that are reporting this week, what are you most concerned about and what do we need to be looking for in the earnings context?

10:38Well, you know, how are things looking in the future? I don't really even care about how things looked for last quarter, I have to tell you. I'm looking for the next six months. And I think especially on more consumer leaning things like Apple, that might be, you know, the focal point for all investors is what new products are driving consumers to the company and are they meeting demand? That whole thing about how in demand are consumer items. Microsoft has a little special issue. I think there's a lot of concern that AI, maybe Claude, is going to replace a lot of software packages that make businesses run.

11:25I don't know how true that is, but I think we're going to get a glimpse into how popular or whether or not businesses are paying up for enterprise licenses, and that's always a concern. Kim Forrest, CIO, founder of Boca Capital Partners. Thanks for joining us across the spectrum of earnings and semis today. Coming up, we're going to the world of quantum. IonQ CEO Niccolò Di Masi is going to be joining us. How his company has just bought yet another quite the acquisition spree. We talk about why it wants to get into semiconductors. This is Bloomberg Tech.

12:12IronQ, well, it's agreed to buy U.S.-based chip foundry Skywater technology for around$1.8 billion. Shares on the downside for IronQ now, they have been up earlier. We're seeing Skywater up more than 5.8 percent. It's a cash and stock deal, and it's the largest in quite the string of recent acquisitions by the quantum computing firm INQ. Let's talk to its CEO now, Niccolò DiMassi. What does having a U.S.-based chip foundry add to the speed of your own technology? It actually does three exciting things for INQ. The first is that it accelerates our entire roadmap. So we talked this morning on a public webinar sort of earnings call style about the fact that it's going to bring in the cycles of R &D design and packaging iteration we do to build our quantum computing chips and hence our full fault tolerant machines quite considerably.

13:02We're moving things in by about a year on our 2 million cubic chip. And we're moving full fault tolerant chips forward into 2028, where we'll have them ready for scale up by the end of the year. What's exciting also is this is a big bet for us on what's called the merchant supplier play in the quantum industry. So INQ already sells our atomic clocks to a number of quantum computing companies in the US. This is going to expand our ability to supply the quantum industry. Our photonic interconnect quantum networking business is also built to cater for multiple quantum computing paths and modalities.

13:44And last but not least, of course, it's a big spread bet on the U.S. semiconductor manufacturing tailwind that I think is already being experienced for critical industries. as we rethink through supply chains, security, and the importance of keeping everything domestic to ensure security. Does that help build and leverage the relationships you already have with the U.S. government, for example? Is that an area that you think is going to drive growth? Oh, for sure. The U.S. government is a very important partner, and the Department of War, of course, is very important. And Skywater obviously does a lot of work, not only for the quantum industry as a whole, where I would argue they are far and away the leading, if not only, quantum foundry.

14:32But Skywater also does work more broadly in classified programs, as does INQ. And so there's plenty of synergies on that vector alone. We look forward to doing more for our nation's government, classified and not. We look forward to doing more for our nation's Fortune 500 companies. And this is a relationship we're really excited about because it increases the velocity and the frequency with which we can spin wafers across all of our product generations, which means the further out we're talking, the more room there is to move things forward. And you're seeing us continue to spend money to move things to the left on our roadmap and win this geopolitical space race of our era.

15:18Nicola, continuing to spend money, there's a lot of emphasis that this is Skywater with US-only operations. Do you look at Europe and other jurisdictions and give yourself backing to go out and do another deal if you need to have a similar footprint and presence in a different market or geography? Well, Skywater has a tremendous amount of upside capacity that we can tap into here, not just on the design engineering side, but on actual manufacturing. So we did this deal because we believe that manufacturability is paramount. Our ambition, as I think you both know, is to be both the NVIDIA and the Cisco, if you will, of the quantum computing and quantum networking and security space.

16:01And so our full quantum platform, networking, sensing, security and computing all has foundry requirements, all has semiconductor manufacturing requirements. What this deal does is it improves certainty, brings forward the roadmap. and, of course, make sure that we have industry-leading costs. So I think this is going to be enough capacity for global production. Skywater focuses on high-security classified program work and U.S. government work, and that's usually a standard that's accepted very internationally for the Five Eyes, NATO, and our allies. Nicolo, in the global quantum race, how do you stack the U.S.

16:41as a leader in field against the prospects of Europe and what you see being worked on in Asia? Well, I fancy INQ's chances, you know, most and first of all. I think that beyond INQ, there's a very vibrant ecosystem in the US. And of course, there is in Europe. The geopolitical race is, I think, very much between the US and our allies and what's going on in China, to be honest. And this is, of course, of paramount strategic importance, not just on the computing side, but also on the quantum security side. These are all solutions that INQ provides. We're the only quantum platform company in history.

17:22And as you both know, we already have a strong foothold in Oxford, England, Geneva, Switzerland, and Seoul, Korea, with our offices that continue to expand to ensure that we're able to not just deliver those solutions, but deliver them in the right global and local way across government and commercial needs in every one of our allies. Well, that was the other big deal that you did, Oxford Ionics, quantum chip technology that you purchased. And I'm interested in how the relationships, though, going back to Skywater are going to continue, because they also serve other quantum competitors of yours.

17:54Will they continue to do so? Absolutely. We are steadfastly dedicated to ensure that the quantum industry prevails. We respect everyone's pathways to market. At the same time, as I mentioned a couple of minutes ago, INQ already is a merchant supplier on our own. We supply atomic clocks and photonic interconnects to the industry and networking equipment. And so Skywater will be taking our existing merchant supplying business and accentuating, deepening and broadening that with, of course, their chips business. So rest assured, we see this very much as a bet on U.S. quantum, our allies' quantum, the quantum industry's supply chain needs, fulfilling those securely and at scale.

18:37And at the same time, in a walled off manner, INQ's own roadmap, of course, will also be accelerated and a supply chain domestically assured. Nicolai, very quickly, you said that this brings things forward by a year. There are people that just aren't familiar with quantum computing. How real is this going to be in the real world and when? Well, we've been saying since we completed the acquisition of Oxford Ionics that full fault tolerant gate model quantum computing is going to happen by the end of this administration. This transaction obviously accentuates that, underwrites that and assures that.

19:18And so I think you're going to see, you know, in the next three years, everything that we have promised on our roadmap come good and hopefully then some. The reason why this is so exciting for us is that with Skywater, we can make quantum computing a mass market reality. And we can do that at a price point that's accessible, not just for governments, but also Fortune 500 companies. And that's very much our goal to build the winning quantum ecosystem across all solutions and product family offerings that INQ have. Security sensing, networking and computing. INQ CEO Nicola DiMasi, it's great to have you back on the show.

19:56Thank you very much. The pressure is building on Grok, XAI and Elon Musk's AI chatbot. The European Union says it's probing X over concerns that it didn't prevent its chatbot from generating deep fake images that, quote, may amount to child sexual abuse material. Bloomberg's London based tech editor, Olivia Solon, joins us. Olivia, what's the need to know here from the EU probe side of this? Well, it's another day, another investigation into X from the EU. This time, yeah, it was about whether or not they did, X did a sufficient risk assessment to prevent harmful outputs from Grok before deploying it across the EU.

20:37This is a legal requirement under the Digital Services Act, which is a kind of set of rules around illegal and harmful content. And so, yeah, there's quite a long list of investigations of X at the moment within the EU, but this certainly adds to it. I mean, you mentioned France has also been looking into it, Ofcom in the UK, where you sit, India. But, Olivia, X's response is they've taken the necessary steps in some way. Of course, it's a unit of XAI. And this, again, builds tensions between the US and the EU in some government manner. Right. I mean, this has been a bit of a flashpoint in the sort of debate over free speech and control over big tech between DC and Europe.

21:24Clearly, I've been kind of checking to see if there's been a response so far from either Elon Musk or J.D. Vance or Trump on this particular announcement. So far, I haven't seen anything, but I'm sure it'll be something that will come up in the coming days. Yeah, the Trump administration has been very critical of this, of the EU in trying to police content on these platforms. And sees it as a free speech violation. We just showed X's statement, but it was a previous statement on how they go about removing illegal content, including child sexual abuse material. Very quick, 15 seconds of it. Does the EU probe have teeth?

22:03It does. I mean, they can find as much as 6 percent of global annual revenue if it's found to be in violation of the DSA. And already they are on the heels of a separate 120 million euro EU penalty coming X's way. previously. Bloomberg's Olivia Solon breaking all down. We thank you so much. Coming up, defence stops. They're on the higher side. In fact, they've soared as governments ramp up spending on heightened geopolitical tensions. More on that next. This is Bloomberg Tech.

22:50Welcome back to Bloomberg Tech. We're halfway through the show and it's a big week in the world of technology when it comes to earnings. Tesla, which reports on Wednesday, is currently the laggard down 2%, but it's a bit of a calendar to track. Now, Apple's the most interesting, up almost 3%. Actually, it's on track for its biggest jump since October. People have been passing the third-party data. There are lots of analysts out there that have said buy the recent dip because it might be better in China than we first thought when it comes to Apple. Of course, you move past that moment in time, Carrie, that is earnings, particularly the quarter gone, and people want to know what's happening with Siri, what's happening with the AI version of Siri.

23:28Mark Gurman's done a lot of reporting on that, of course, but until the company says something material, that's still what the hopes and expectations of the street rest in. What are you looking at? It's an alphabet story too, isn't it? But I'm going to pivot away from the earnings story just for a moment Ed because we've got to talk about geopolitical tensions. We have seen the effect it's had on defense stocks. They've soared as you'll see. In fact the sector gained nearly 44 % you've seen in the last one year basis. Military contractors are cashing in as governments to ramp up the spending on fighter jets, yes on missiles but also tech areas, drones and some investors really say that the rally is just getting started.

24:05Here with more is Bloomberg Equities reporter, Avelisse Bonilla-Ramos. And Avelisse, is this sustainable? What are we expecting in terms of revenue growth, profitable growth, fundamental growth on these earnings? Sure, Caroline. Thanks so much for having me. And yes, political developments are definitely a major driver of this sentiment. You know, the rally is driven by expectations that the sustained increases and global defense spending, this geopolitical instability, shifting alliances, ongoing conflicts are going to sort of like, you know, boost more this rally and continue this rally. And then these earnings this week are going to be really a test to what can actually happen.

24:49And, you know, the Aerospace and Defense Stocks Index is rising or rose about 42 % last year, you know, pacing a little bit of the broader market. Analysts are concerned about certain things you know these stocks are um in the political crosshairs just to put it that way um definitely the well the other concern is yeah sorry let's let's do a bit more on the political crosshairs when caroline and i were in las vegas for ces uh the president on true social was very clear ahead of an executive order that he wasn't happy about uh defense company compensation the compensation of the CEOs at the head of these primes because they weren't delivering.

25:31That seems incongruous with the massive run up in those stocks. Sure. I mean, the investor confidence right now, and that's very interesting, but the investor confidence right now in Momentum is sort of like these companies are growing a little bit faster than what the sort of like classic thought about or behind this defense talks usually is or was because a lot of these companies are moving into the technology play. There's a lot, you know, modern warfare is increasing. There's relying on more reliance on drones, on AI, on advanced sensors, software, missile defense systems. So once all these companies are sort of like tapping into that, investors continue to be a little bit more positive on this side.

26:19And that is exactly why we have you on, because of the tech focus that these businesses are having. We'll watch out for their earnings. Bloomberg's Aveliz, Bonilla, Ramos, thank you so much for joining us. Meanwhile, let's turn to wintry weather. It is blanketed. Most of the United States is leaving lingering effects, too. Power grids, they're expected to endure increased stress from really brutal cold in the wake of the damaging winter storms. Get this, more than 800 ,000 homes and businesses nationwide are currently without power. And this is concerns already have been growing around the U.S.'s power grid's resilience in the face of rising AI power demands.

26:49The first power and natural gas reporter Noreen Malik has been very busy this weekend reporting all of this out. What's the effect on prices? Yeah, prices got really high yesterday. You know, weekend is usually a lull for prices. And we saw prices around like a thousand dollars this morning. The Texas grid in like West Texas, where they're usually lower, was like two thousand dollars. And in Dominion, in northern Virginia, where there is that concentration of data centers, we got to like$1 ,000 as well. Spot prices are a little lower because real-time demand seems to be affected by school closures and businesses, just everyone working from home.

Read the full transcript

27:25But it's going to be an extended cold, and that's what's most concerning right now. Noreen, out here on the West Coast, we're used to a very similar story, but the opposite, right? in the summertime, extreme heat, the grid being under pressure, and either the utility can or can't cope. What is it like from a technology perspective or just operations perspective that the grid can actually do to manage and cope in scenarios like the extreme cold we've seen and weather on the East Coast? Yeah, I think what's interesting about the East Coast and even down to Texas is they've also become winter peaking grids, not just summer like California.

28:02And so they're basically trying to take out all stops. You've had the Energy Department grant authority to Texas, New England and PJM, which is the largest grid that stretches from the Mid-Atlantic to the Midwest. They've granted them authority to run power plants basically all out, irrespective of like emissions limits. The DOE also told ERCOT, the Texas grid operator, that they can use backup generators at data centers and factories to supply the grid. And last night, PJM had a call with members to tell them that, hey, we're putting in place measures to also do the same, to use data centers in lieu of blackouts if things got that bad.

28:42Right now, they're expecting to have enough power. Who bears the price? Who bears the cost right now? Ultimately, the consumers. Everything is passed on to consumers. How quickly will depend on your utility rate. But wholesale prices are often a pass-through cost. noreen malik uh bloomberg busy weekend for you across what is an astonishing visual story as well thank you very much now coming up base 10 ceo tohin shri bestava and one of its backers conviction sarah guard they join us as the startup more than doubles its valuation we have the details next confirmation of that round this is bloomberg tech

29:39AI inference startup Base 10 Labs has raised$300 million. The funding round lifts Base 10's valuation to$5 billion, more than doubling what it was just six months ago. Base 10 CEO, Tuhin Srivastava, and Conviction founder and partner, Sarah Guar, who participated in that round, join us here in San Francisco. I've been using a number of generative AI tools this morning to try and distill down what it is that Base 10 does. I'm going to read you one, and you can tell me, and I won't say what engine it was. Base 10 provides the software and infrastructure that lets companies run AI models reliably at massive scale only once they're built.

30:14It's about being fast, stable, and cost efficient. Is that what you do? And why did you need to raise$300 million to do it? Yeah, well, the way we think about it is that inference is probably the largest market that will ever exist. I think you see this amazing set of companies like Cursor or Bridge, Open Evidence, Notion kind of growing with their inference workloads and to support the scale of these massive companies and you know we're just arming ourselves there more importantly the amount of compute that we're going to need to support this is going to be very large and again compute is a capital intensive business and that's the business we're going after lots of very interesting participants in the round and Sarah it's great to have you back on Bloomberg Tech I was really interested to see conviction in there right you know the size of the round the valuation's interesting, but you're all about helping founders build up, you know, at Conviction.

31:09Why did you want to do this? Well, I'd say this is a very secular bet for us. First, we have invested in every round of base 10 since inception from my prior firm and then through Conviction. And so I think five billion actually is a very small number. Just say Greylock was your prior firm. And then at Conviction, my current firm, five billion dollars is a very small number compared to the opportunity ahead. I genuinely believe that. And so, you know, we're looking at the difference when we make that investment. But I think from a broader perspective, NVIDIA and Base 10 and Conviction, we have the same vision for what needs to happen to enable all of these AI native startups.

31:46Inference engineering, RL as a service, the optimization of models to make them cheap and efficient to run. These are like the three hottest jobs in AI right now. And one thing Base 10 does is take these capabilities and use NVIDIA capacity and make them available to everyone rather than just the labs. Let's talk there for a little bit, Tuhin, about you said you're arming yourself for this opportunity in inference. NVIDIA is interested in investing, but how much do you end up using technology, competing against? Where do you see the future of what it is that you're solving for? Yeah, look, we've been partnered with NVIDIA for almost five years now.

32:29They're a big part of the future. We work really, really closely with the chips they provide. We don't see them as a competition. Honestly, we think everything's kind of a rising tide, and we want to work with the biggest and best players and customers and suppliers and vendors and technologies, and NVIDIA is a big part of that. Sarah, to that point, it's interesting the amount of news we've had around NVIDIA today, whether or not it's another what people call circular financing deal, backing CoreWeave, buying stock, indeed talking about a CPU, and indeed we see Microsoft announcing yet another chip AI version for themselves too.

33:07How does this bear out? Does everyone win? Are there going to be select winners? Well, I think so far it looks like NVIDIA is a pretty strong winner at the chip level. I do think that there are enormous investments in accelerators, particularly at DeepMind, but there are other legitimate efforts as well. And I think it all speaks to the overall demand in this market. The reason people are making these really strong investments in other chips is because they're going to go spend billions and tens of billions and hundreds of billions of dollars, trillions, if you listen to Sam, against this compute over the next decade.

33:43And so I think the question of winners is the problem today is just meeting the demand. and all of these players are trying to come up with alternatives, less about cost and more about creating more supply. I enjoyed reading about and learning about Base 10. There's a section on your website where you break down pricing, and I don't know, that might seem really mundane to both of you, but you basically have pay-as-you-go at the low end right through to enterprise-level deals. Explain that kind of strategy. You know, we've discussed with Cursor the idea of per desk versus pay-as-you-go. It's actually a very interesting model.

34:21Well, I think pay-as-you-go pricing fully aligns ourselves with our customers. So we are only getting... Who are they? Who are our customers? So think about the fastest-growing AI companies in the world. So anywhere from Abridge, Open Evidence, Cursor, NOSHA, Gamma, Clay, all these are companies that you've covered, I'm sure, over the last year. These companies are growing incredibly fast, and they are using a lot of inference. And by aligning ourselves on a pay-as-you-go model, we only make money when they are getting value out of the models that they are training for their customers. And so that's actually a very important part of our business.

34:57We think it's a very important part of the ecosystem. And it's all about being a long-term partner to the best companies in the world. And Sarah, what gives you the confidence that this doesn't get commoditized or at least how sticky are the customers when it comes to inference? uh i i think training is a job by job sort of buy right inference has the stickiest characteristics of any business i've ever seen right and so you know without giving specifics the ndr on base 10 is through the roof and we see customers they come with us they scale with us um they don't want to take on this enormous engineering burden of you know these different specializations i described.

35:38The labs want the kernel engineers too. They're very expensive. Managing across 10 different clouds is very expensive. Owning your own capacity. Ed was just talking with Tuhan about the pay-as-you-go versus the CapEx upfront model. The application winners, they don't want that. They want to go after new business and creating value for users. There's a lot of energy in your industry still. It's really cool to have you both here in person. Deal flow hasn't slowed down at all. I wondered what the talent story is right now. Sarah was talking about the three kind of core roles, hottest jobs. I'm assuming that a chunk of that 300 million reflects the capital intensity of talent right now as well.

36:17Yeah, look, we've been very lucky to be able to attract really great talent up until now. That being said, as Sarah said earlier, the folks that we are going after are very, very in demand. If you think about inference being this massive market and then this big systems engineering and kernel engineering problem that needs to be solved. And probably there's, you know, sub-1 ,000 people in the world who have solved this at scale. Yes, it is going to be very aggressive and, you know, to win the ultimate prize for us, which is, you know, being the inference cloud, the AWS for inference. It's a big target to have.

36:52It's a big target to have, and, yeah, we need to be able to get that talent. Sarah, I also know you as kind of Sarah the operator, not necessarily just writing checks. We've been talking about is circular financing a bad word, for example, with NVIDIA and CoreWeave. But, like, is there an opportunity here for you to look at the rest of your portfolio and your historic investments and say to some of these companies, like, you guys should work together? What's the reality of that? Well, I'd say Base10 does have a lot of customers they've won within the portfolio. And one of the things that we do at small scale that's very related to this NVIDIA move of enabling its ecosystem and its customers is have an incubator.

37:27Base10 is a contributor to that. They give credits to startups. They're offering their expertise to these companies. We give people compute upfront as well. And it's to reduce the barrier to entry for the next generation of application companies. This program is called Embed. And I think our entire investment is we want to see more of these AI native application companies win, which Base 10 does and NVIDIA does too. To Hindshir Vistava, Base 10 CEO, Conviction Founder, Sarah Gawar, it's great to have you both here in San Francisco. Thank you very much. Caro, there are so many more tech headlines out there today.

38:00Boy, are there. And around compute and such. We're talking SoftBank first. It stopped talks to buy data center operator Switch. Now, that's according to sources. The company has been pursuing Switch for months, offering as much as$50 billion. But the deal would have helped build out SoftBank's Stargate AI infrastructure. Discussions continue, though, for a partial investment or a partnership. Plus, Samsung is getting close to securing a certification from NVIDIA for its latest AI memory chip, the HBM4. Sources say the South Korean company has entered the final qualification phase after supplying samples to the U.S.

38:30chipmaker in September. Samsung is said to be preparing mass production for the HBM4 in February. And shares of USA Rare Earth, they're surging. The company entered a non-binding letter of intent with the United States Commerce Department for$1.6 billion in funding in exchange for 16.1 million shares and warrants. Now, the company also raised$1.5 billion in private investment to accelerate domestic heavy rare earth production.

39:03From Wall Street to rural America, today's AI is the economy. And markets have priced it like a miracle that can't fail. Investors are really banking on incredible growth. Microsoft, Meta, Alphabet, they are just committing billions and billions of dollars to capital expenditures. But in this burgeoning industry, what goes around doesn't necessarily come around. A precarious investment strategy is emerging. Multi-billion dollar circular deals. Huge, huge sums of money are just being passed between these enormous companies that are pouring hundreds of billions of dollars into the promise of AI.

39:44And the promise is huge. So this is a structural shift, like electricity or the internet. But AI, with all its potential, remains largely unproven for profitability. Probably the biggest question in San Francisco right now is, are we in an AI bubble? And if we are, then how big is this bubble and how bad would it be if and when it does burst? So, is this the dawn of a new age of AI-powered growth? Or the biggest bubble ever?

40:19You can watch the full episode of that Bloomberg Originals piece on YouTube and Bloomberg.com. We have another alleged example of circular financing just today. NVIDIA investing another additional$2 billion into CoreWeave. NVIDIA CEO Jensen Wong saying that it is ridiculous to call it circular financing because it is just a tiny portion of what CoreWeave is going to have to raise overall. And it's in line with NVIDIA's other investments. Make your own mind up on that. There's also a piece on the Bloomberg about how NVIDIA's rally of late shows that a year ago, deep seek fears might be a little unfounded.

40:50And that is written by Bloomberg's Carmen Reinecke. Carmen, great piece, as always, data focused. What do we need to know here? What are you writing about? Yeah, so I think the thing that's most important or interesting to watch here is how, you know, a year ago, this deep seek news came out and NVIDIA sold off, you know, almost$600 billion in market value was erased in one day. And it looked like maybe the startup or maybe a practice run for the AI bubble popping. But if you look a year on now, a lot of those fears are really unfounded. You know, the stock is up nearly 60 % since then. We've seen it just continue to grow revenue.

41:26And overall, the AI ecosystem has also continued to develop. You know, a year ago, NVIDIA was still by and far the kingmaker. It was seen as sort of the only show in town when it came to chips. And now we're just seeing so many more businesses sort of step up, come to the plate with their own chips, trying to not be so reliant. And then, you know, we're also seeing, you know, more energy and construction and infrastructure stocks kind of come online. So while there still are concerns about some of the circularity here for sure, and even, you know, the deal with Core Web and NVIDIA goes deeper than just the most recent news, the overall sort of expansion of the AI trade is giving bulls sort of more room to run.

42:07What do we need to see this week, Carmen? For example, Microsoft today unveils that it's got yet its second round of its own AI accelerator pointed maybe at trying to reduce its dependence on an NVIDIA. But we're also going to see what its capital expenditure looks like, how much it's continuing to invest. That's going to be important for NVIDIA's numbers that don't come from until the end of February. Exactly. Really, the biggest thing I think that investors are watching in these big tech earnings that we're seeing over the next few weeks is that CapEx line item. Are these companies continuing to grow or say that they're going to spend even more?

42:41That's really sort of paramount to the rest of the AI trade or how it continues to develop. And then on the flip side, I think investors will also be looking for signs of ROI, right? We want to know that all of this money is going to something that it's going to help the bottom line, that it's helping efficiency and to drive business forward. So really two big things that we're going to be watching. Bloomberg's coming, Reineke. It's great to have you on. It's a great piece that you also are in on Bloomberg Originals. Meanwhile, that does it for this edition of Bloomberg Tech, Ed. Yeah, strong start to the week.

43:15Caro has made it through the storm to the studio and there's a lot more to come. recap on the podcast you know where to find it on the bloomberg platforms apple spotify and iheart online stick with us this week this is bloomberg

From the publisher

Bloomberg’s Caroline Hyde and Ed Ludlow discuss Nvidia’s additional $2 billion investment in CoreWeave, to accelerate the buildout of AI factories. Plus, IonQ CEO Niccolo de Masi discusses the quantum company’s continued acquisition spree, as it buys chipmaker SkyWater Technology in a cash-and-stock deal worth $1.8 billion. And investors gear up for big tech earnings with Meta, Apple and Tesla all set to report this week.

See omnystudio.com/listener for privacy information.

More from Bloomberg Tech

All 343 episodes
Nvidia Puts Another $2B Into CoreWeave, Offers New ChipsBloomberg Tech · 44 min
Listen in VO