Nvidia’s Retreat of AWS-like Cloud Ambitions, SpaceX IPO and 2026 IPO Outlook | Dec 23, 2025

0:13Welcome, everyone, to the information's TI-TV. My name is Akash Pastracha. It is Tuesday, December 23rd. We are kicking off the show with ServiceNow's big acquisition of Armus. We have a top ServiceNow executive coming on the show momentarily to talk about that deal. We'll then get to our story about NVIDIA reorganizing its cloud division, a shift that pulls back from Jensen Huang's earlier ambitions to build a cloud service that could rival AWS. We're then looking at the IPO landscape. Wall Wall Street entered 2025 with high hopes, but several high-profile listings have stumbled out of the gate, just like I just did.

0:53We'll dig into what's going on. And we will wrap the show with a conversation around SpaceX's possible 2026 IPO and how Elon Musk's plans to sell it to Wall Street. The CEO of StarCloud joins me for that conversation. It's going to be a fun episode. There is a lot going on. We are two days from Christmas. Let's get right on into things. ServiceNow is buying Armis, a cybersecurity company, for$7.75 billion. It has been a busy year for M &A for ServiceNow. And I want to bring on Amit Zaveri, the chief product and operating officer, to talk to us more about why the company is doing this deal. Amit, welcome to TITV.

1:32It's great to have you here. Thanks for having me, Akash. Great to see you. So you guys are busy right up until the holidays here. It's been a busy time, but as you see, the market has been very busy as well, right? So the industry is moving very, very fast. A lot of new requirements coming from our customers. And customers are also grappling with a lot of change in technology landscape. So we have to make sure we stay ahead of it and ensure that we deliver what they require. Well, I want to talk to you about the deal that you announced this morning. So look, Armis is doing$340 million in ARR.

2:05You're paying more than 20 times that amount for this cyber company, which is a higher multiple than a lot of public cyber companies right now that are trading. What justifies the price and why did you decide to do the deal? Yeah, I think if you look at what's going on in the cyberspace, customers are grappling with huge amount of changes happening with amount of machines out there, amount of AI agents, as well as the amount of security threats they're dealing with. The complexity of making that manageable is becoming very difficult for customers, right? So Armis has been kind of the forefront of how you do exposure management.

2:43They have a lot of depth in terms of understanding vulnerability, as well as any issues customers need to manage proactively before those security threats become real. Armis is a very fast-growing company. They're growing at 50-plus percent, as you mentioned, at$340 million in ARR. And they have been also innovating very fast. They've been creating the full end-to-end capabilities around machine data, things around IoT devices, operational technology, or things on the shop floor, as well as AI agents. And they're bringing all that stuff together into one integrated platform, which works very well with ServiceNow.

3:20We've been a partner with them for many, many years, and we continue to integrate that and bring that value to customers. Seven out of top 10 Fortune companies work and use Armist today. Companies like United, Merck, and others have been really depending on Armist as the way to manage their cybersecurity threats. So we see a lot of value creation jointly as we can take this to market. So everyone who knows ServiceNow knows that you guys are the organic growth machine. I mean, for a while there, it seemed like ServiceNow was not doing any big acquisitions, really. And now lately, we've seen a spate of these large deals.

4:00Why all the M &A? Yeah, I think we have been also, and I think we're not dependent on M &A for our growth. So what growth we have been telling to Wall Street, around 20 % growth. We are probably the only company with 10 years plus of rule of 50, delivering 20 plus percent growth and 30 plus percent of free cash flow margin. And that will continue. So none of that stuff is changing. That remains a trajectory and a plan. And we're completely confident about delivering those numbers. But I think there's an opportunity to be had in the market. Can we even accelerate that further? And what are the areas where we see a lot of value creation for ServiceNow, a way we can accelerate our roadmap and bring in capabilities, which we think can really add more value to our customers.

4:44So with Armist, I mean, you look at our security business already, which is billion-dollar-plus in revenue, which is built organically. I think we can accelerate that even further with capabilities Armist brings to our portfolio, as well as bringing in some of the domain expertise we require as we keep on building out new capabilities as well. So our organic plan remains intact while we add other things around it. I want to come back to that risk and security business. It's the fifth business above a billion dollars in ACV now. But very quickly, I mean, before we talk about other potential lines of business, I am curious in cybersecurity right now with AI making it so much easier to code anything, whether it's coding something for good or coding something for bad.

5:31I mean, there is kind of this question in my mind around is AI net good or net bad for cybersecurity? because, I mean, you could make the argument both ways. No, you're right. I think cybersecurity is getting really threatened by AI, right? I mean, there's a lot of issues where the companies who are trying to deal with AI, they don't know what to protect, how to protect. They are adding a lot of new features and technologies into the existing technology stack. But how to protect that is becoming a big challenge. That's why we introduced something called AI Control Tower to help that and give visibility as well as control to our customers.

6:11But that's not enough, right? You need to have a lot of intelligence built in and the threat actors can use AI in a different bad ways, as you mentioned. So I think to remain proactive and vigilant, you have to take on a lot of the AI capabilities into your platform and make sure you're making that protection being provided to every customer out there. So it's going to be a battle for some time. And if you don't invest now, you will be left behind. And our view is that we need to be very proactive and invest in the new opportunities we see, as well as customers asking us as well. Seesaw is a big buying center for ServiceNow.

6:49And as you rightly pointed out, it's one of the largest and fastest growing businesses for us. And we see a reason for that, because they'll be trying to solve a very complicated problem customer face today. And we can be the leader in that area for that. So let's get back to sort of the broader view of the business here. You have five of these businesses underneath your product umbrella now. You've got businesses in IT, which was the core of ServiceNow, obviously. You have the HR workflows and the customer service management, CRM workflows. You have cybersecurity now. So you got these five businesses.

7:28What are you gearing for to be the sixth business above 1 billion in ACV? Yeah, I think we see the way we're seeing a lot of new traction now is around data analytics, right? So that's a space which is becoming quite important because that's the foundation for how we do insight to action. Given that we have visibility end-to-end across a customer landscape and we're the enterprise OS, we have a very good understanding of what's going on in every business process, how to kind of manage that, how to really make that thing more efficient, but also how do you get insights so you can run your business with a lot more information and make the right decisions.

8:04Given that we run this at a business process level, we're seeing a lot of value creation happening around the data analytics stack as well. So our goal is to really continue growing all the businesses we have and make them multi-billion dollars while also incubating new areas as well.

8:23How far are we from data and analytics becoming a$1 billion business? We don't share that number publicly yet, but I think the goal for us is that we have to see where our customers are taking us, what we see as the evolution of our platform stack. And our goal always has been connecting everything east to west and north to south across an enterprise landscape. And that's where the differentiation for us has always been. And that's where the value creation. So data is a natural extension. We have announced a lot of new capabilities around that with the Raptor DB, workflow data fabric, a lot of the work we do from insight to action and BI apps associated with that.

9:00And that plays very well in our IT landscape, what we've been doing with the tech and the core part of our business. Security plays also in our core IT landscape as well. So what we're doing with OT and IT, that thing combination is also very, very unique and very differentiated as well. So our goal is always to kind of look at where our strengths are and how we can surround that and can create a full capability for customers so they don't have to do piecemeal, don't have to buy many, many different parts of the products, and then have to cobble them together and different systems which don't work together.

9:29So we've always been a believer of any system, any data, any cloud, any AI, and making it all work together for our customers so they get a lot of good value directly from ServiceNow. Great. Well, Amit, I want to thank you for coming on. Congratulations on the acquisition. Hopefully you find some time to rest over the holidays. That is Amit Zaveri, Chief Product and Operating Officer of ServiceNow here on TITV. Okay, NVIDIA is stepping away from competing with cloud providers like Amazon Web Services and has reassigned some of the senior executives that were working on that effort. That is according to a scoop that the information was first to report.

10:08I want to bring on Wayne Ma, our NVIDIA reporter who co-authored that piece, to tell us a bit about what the team was supposed to have done and why the company decided to pivot. Wayne, welcome back to the show. It's great to have you here. Hey, Akas. So let's talk about NVIDIA's cloud business. I mean, this is something that we've talked with Anissa Gardizi on the show about. She co-authored the piece with you. What was the cloud division? I know these three-letter acronyms, DGX, what do they stand for? Tell us about it. Well, initially what NVIDIA wanted to do was create a cloud service that was heavily optimized for its GPUs.

10:51So there's lots of cloud providers that buy its GPUs and put them in their own servers, but they have their own configurations. And so NVIDIA wanted to set the bar really high and showcase kind of their hardware in their own type of servers. But then they needed a cloud service to run that. And so they kind of built this cloud of clouds where they built a cloud service on top of other cloud providers. Okay. And how much traction did it get? Not very much. So they wrote that the business could potentially be, you know, worth$150 billion in revenue in the coming years, which would be more than what AWS generates annually.

11:26But in reality, they didn't have very many customers for the service. It wasn't very reliable. It was hard to support because it was running on multiple different cloud providers. And so eventually, recently, they decided to kind of retreat from that operation. So you say relying on multiple different cloud providers. So NVIDIA's cloud business was depending on the other large hyperscaler cloud businesses? Is that the idea? Right. So they wouldn't build their own data centers or operate their own data centers. they least capacity from existing hyperscalers. Okay. And so was that the only challenge or what were some of the other challenges that made it so difficult to scale this thing?

12:08Well, I think that at the time they announced the service, their chips were in short supply and lots of these hyperscalers were withholding them from smaller startups and reserving them for larger customers. So NVIDIA wanted to kind of break that. But in the period between when it started the service And now there's lots of smaller cloud providers came to market funded by private equity or other types of money. And so the service wasn't actually needed as much as they thought it would be. And how does this relate to Leptin? This was another category of its business that we've talked a little bit of on the show.

12:43Well, so then instead of creating their own cloud service that they set up customers for, they're creating like a marketplace where the hardware that they were supposed to use for that is now listed on the marketplace and it connects basically those cloud providers with customers. So they're not effectively in the middle anymore. Got it, got it. I wonder what this decision tells you a little bit about NVIDIA's broader strategy right now. I mean, it sounds like they were trying to become something much larger than the largest chip maker or chip designer, I guess, in the world. it seems to have narrowed its, if you can say that, it's narrowing its ambitions a little bit.

13:24What does this tell you a little bit about Jensen Huang's broader strategy right now? Well, I think they're trying lots of different things to diversify their business away from just selling hardware. And so, you know, 90 % of the revenue comes right now from selling AI chips to data centers. And I think they were hoping that by offering a cloud service, maybe that business could grow and be another source of major revenue for them. Do you think it could ever come back? I don't think so. I don't think the company really has the expertise required to run a cloud service. It's also not that profitable compared with selling ships.

13:56And so I don't see them reviving this business anytime soon. And you also wrote about where the executives who worked on this effort are going. What are they working on? Well, the lead, the head of the division, she's currently planning to take a new role in the new year. It hasn't been really decided yet. and the rest of the group has been folded into the engineering and operations divisions. Got it. And so heading into 2026, Wayne, for you following this company, what are the big questions that come to mind for you as a reporter, areas that you're really curious to drill deeper into for the company?

14:34Well, I think that the big thing is how are they signing up these deals and signing up these large customers in ways that use creative financing. So one of the big issues is their chips are so expensive that NVIDIA is actually trying to help its customers finance them in these very weird ways, like creating special purpose vehicles to put the debt in, exchanging the chips for equity, like as you can see with OpenAI. So I think in the coming year, what I'm trying to follow is more how are they funding these chips? How are these chips getting funded? Right. Well, Wayne, I want to thank you for coming on.

15:10That is Wayne Ma, our NVIDIA reporter, here at The Information. Okay, everybody thought that 2025 was going to be the big year for IPOs, and in some cases, it certainly was. We finally saw some big names go public like Figma and Circle and CoreWeave and Klarna, but you do not have to look far to see that of all the tech companies that have gone public this year, only a select few are trading above their IPO price, which is casting doubt on the many other big tech companies waiting in the wings. Joining me now is Corey Weinberg, our Deputy Bureau Chief of Finance, to tell us more about what he's hearing on this beat.

15:51Corey, welcome back to the show. It's great to see you. Hey, Kof. Great to see you. Well, I'm excited to talk about our favorite topic to close out the year, which is the year of IPOs that we've had and the year of IPOs that 2026 could be. So you published a story on this today, and my read on it is what? SpaceX is ready to go, but everybody else is kind of not? I mean, what's the deal? Yeah, I think we have this dynamic in the IPO market right now where investors seemingly have been honestly like not that interested or that enthusiastic about the types of companies that make up the vast majority of the VC backed tech IPO market.

16:39this kind of crucial segment that's often been sort of a key engine of the IPO market. These companies that are valued in sort of the$5 to$15 billion range, doing hundreds of millions of dollars of revenue, maybe close to profitable, a lot of them have met a fairly cool reception after their debuts. These are companies that are pretty notable in the VC-backed world. Klarna, Navon, Chime, StubHub, companies that have well-known brands, they've traded poorly. Companies that are the sort of mag-7 of the private markets, if you will, the SpaceXes, the OpenAI, the Anthropics, those are the ones everyone's pining to get a piece of.

17:31You had a name for this group, right? You called it this? Yeah, I called them the shy five. The shy five, right. These are the private companies that were not looking to make their public debuts anytime soon. Right. Okay. Well, I mean, we'll get back to the shy five in a second. But so the companies that have gone public, I mean, you noted in the story, companies like Klarna, StubHub, Gemini, I mean, their stocks haven't exactly done the way I think maybe some people would have hoped out of the gate. Does any of that have to do with the lockup periods? So I think that is sort of the defense that I think a lot of bankers would say is like, especially companies that went public in the September, October period, which was a pretty heavy period for IPOs.

18:17This was after Figma went public and there was kind of a real opening in the market before the government shut down. You know, obviously it hasn't been six months yet. So I think there is a case where investors sort of wait on the sidelines before they buy a stock before the lockup because they want to see how the stock performs because often a stock goes down after sort of a lockup expires because there's more shares that hit the market. So that's certainly a dynamic. But I think there's definitely a more qualitative aspect here rather than just a kind of market driven, you know, sort of technical technicality.

18:52And I think it's just like I hear it from IPO investors all the time. Like people just honestly, like, you know, at a certain price, they care about buying, you know, a StubHub or a Nirvana or a Chime, but they aren't kind of necessarily excited about bidding them up. You know, they have a lot of choices out there in the public market. It's not the same sort of scarcity-driven environment that the private markets are. And so I just want to understand the dynamic here. So when you say the bankers, they point to this lockup period, and then there's also the reality that the lockup period hasn't actually – I mean, it hasn't been six months yet, as you say.

19:28Is the idea here that – I mean, bankers would say, oh, people are not buying the stock right now because they're nervous that when the lockup expires, then they can sort of expect some kind of a further tank? Yeah, I mean, there's a reality there. I think it's sort of overall, I think the takeaway is investors are wanting to wait and see sort of what what happens with these companies in that sort of five to$15 billion market cap range. They want to see how does it perform after the lockup expires? They want to see how does the company more importantly, how does the company perform? You know, are they actually quarter after quarter doing what they said they were going to do?

20:11That's a pretty basic sort of, you know, sort of dynamic of being a public company. But when you don't have the same sort of FOMO-driven, retail-driven backstop that, you know, a company that's in crypto or AI or, you know, a SpaceX would have, you know, you're not going to get the benefit of the doubt because people are not chasing that stock necessarily. Right. So let me ask you this, Corey. You've been covering this space now for a while, and you also have a side of you that you really like to go deep on specific companies. And you've been at the information now for 10 years, we should say.

20:45You just celebrated your 10-year anniversary. And we were taking a trip down memory lane. You covered companies like Airbnb closely. I mean, more recently, there was Scale AI. When you think about the companies out there right now, They don't have to be pre-IPO companies, but what company going into 2026 is just the most fascinating business to you right now that you're watching into the new year? Oh, that's interesting. I mean, given my move to Los Angeles, I definitely am trying to drink, swim in the waters of defense tech a little bit more. And so you have a company in the private markets like Anduril, which I've written about a fair amount over the last year or so, where they're at this really interesting crucible moment where they're valued at$30 billion.

21:34They're generating about$2 billion in revenue. They're benefiting seemingly from a shift in the Pentagon's sort of attitude towards sending contracts to younger, more unproven companies. and this is going to be the company whose IPO which will probably come, I don't know, a year after SpaceX maybe a little bit longer. I don't have exact information on that but it'll be after SpaceX and they are smaller, younger than SpaceX. If the Andrile IPO doesn't necessarily go according to plan that's going to have a huge ripple effect on the rest of the entire defense tech class, so to speak, where there's been a ton of venture money going into it.

22:24And I presume there's a lot of these late-stage defense tech companies popping up now that are sort of rallying around? And very speculative, and they have very little actual revenue. There's a lot of vibe investing going on right now in the tech world. Some of it's obviously legitimate, but you know venture capital. that it's a business where you're rewarded for taking bets that are not sure things, obviously. Great. Well, Corey, I want to thank you for coming on. I look forward to having you on more in the new year to talk about defense tech or IPOs or anything you want, really, because, well, you've covered it all in your 10 years, and I'm excited for the next 10.

23:08Thank you for joining us. That is Corey Weinberg, our Deputy Bureau Chief of Finance, here at The Information. Okay, SpaceX is up for a probable 2026 IPO, as we just talked about. The information was first to report that news exclusively a few weeks ago. But one of the bigger questions right now is the extent to which Elon Musk will be able to sell the company's story to Wall Street and what story might be the most effective pitch to investors. There are a lot of thoughts out there on this. One person who knows the space very well is StarCloud CEO and co-founder Philip Johnston. And I want to bring on Philip to help us understand his thoughts on the matter.

23:50Philip, welcome back to the show. It's great to have you here. Thanks so much for having me back. So we had Tim Ferrara on the show a couple of weeks ago. We were talking about the SpaceX IPO And he made this point. He said, Elon is attaching himself to the data center in space story because it's the next big thing. And this is what SpaceX needs to market itself for a successful IPO. And, you know, you had a tweet that did pretty well, a post on X. You said, you know, it was a meme. Do you want to tell people what the meme was? uh i i think i was not particularly polite about his um his opinion but i don't remember was it i mean yeah anyway it was it was it was the frog of shame okay it was the frog of shame yes yes yes okay i remember now so and look the frog of shame got more likes than the post okay which is the whole point of the frog of shame so so what i want to know is what prompted you to post the frog of shame and tell us a little bit about how you're thinking about it yeah i think that the tweet that gone before that was me saying uh something along the lines of um to say that spacex is doing data centers in space for the story because it's worried about missing its numbers um i think i said was the dumbest thing i'd heard all in a quite a long time um like spacex is not worried about missing its numbers spacex the spacex numbers growth story is one of the most impressive corporate revenue growth stories in history um you know i think they tim tim says tim said something like they were worried because they had forecast 15.5 billion and they're only going to get 15 billion of revenue which to me is just like an absurd thing to say that now they're doing data centers in space because they're worried about half a billion miss on a basically zero revenue in Starlink three years ago to now half the company's revenue.

25:51I mean, it's an incredible growth story of revenue. My view is it's completely opposite. SpaceX has realized that doing data centers in space is the cheapest way to do compute and they want cash for it. And that's why they're IPOing. And I think Elon himself even confirmed basically, maybe you can believe what you want. But my reading is that that's from broadly accurate is that they're trying to get cash to finance building data centers in space because it's pretty capital intensive. Yeah. Now hear me. So, I mean, I take your point, you know, it's a private company. So forecasts are, they're hard to know, you know, what the exact forecasts are.

26:31And, you know, I think one of the points that I took from Tim was that if you do want to be a public company, I mean, you will have to give some guidance or commentary at least to investors and at least know that you're going to be in that range and not consistently fall short. I mean, is that tough to do in the space category? It feels like just because of how burgeoning the space is, I mean, it might be hard to be a public space company in 2026 or even 27. um maybe but um forecasting 15 15.5 and getting 15 is the kind of thing that a company that's been public for like 100 years might do you know if or even you know more established companies if you know if nvidia or apple or anybody else were to do that it wouldn't be like the most crazily outrageous uh miss of all time um so i mean they're getting pretty accurate at forecasting numbers to be honest and it's only going to get more accurate because the starlink revenue stream is like one of the most stable cash um most predictable cash flow streams in the in the entire in the entire world i mean i can't think of a more stable cash flow stream than than that so yeah i think it's gonna i think it's only gonna get even more and more stable so let me ask you this you are and we should remind folks that star cloud i mean you are the first if not one of the first companies to really get some of these chips out into space.

28:04And I saw that you recently put out an announcement that you've been training models in space as well. You know, I think the other part that people might be reacting to from Tim's comments is this idea that Elon Musk has a, you know, he likes to attach himself to bold missions, right? And those missions sometimes can be three years, 13 years. I mean, it's really hard to know when that will actually become a commercializable operation. So AI data centers in space, I mean, you know, what is your timeline on when you think that could be generating consistent revenue for any business, let alone yourself?