Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is alive from coast to coast with Caroline Hide in New York and ever though in San Francisco.
This is Bloomberg Tech coming up. Google's thinking about the next phase of its TPU program and that moves markets will have the latest.
Plus ast space mobile shares plunge after the Blue Origin space launch miss over the weekend. We also spoke the Blue Origin CEO.
And IPOs are so back, Sarah brass IPO is back on the table months after we're drawing a previous attempt.
So first we check in on these markets, which are dictated by geopolitics once more. And look on the day we are down on the Nasdaq one hundred to the tune of seven teens percent. We're actually heading session loans. Why the latest headlines coming from President Trump that the ceasefire that currently is holding the truth that they had put in place for two weeks is unlikely to not be extended. It's unlikely to be pushed forward, even as though we anticipate some talks to kick start once again
over in Pakistan on Wednesday. Now, there's also concerns really about the Horll moves, the straight of horm moves likely to remain blocked ed. What this means is that the streak of tech stock wins is likely to be coming to an end. We're up a eleven point six percent of the thirteen days. If it had held for a fourteenth day of gains, that would have been the longest winning street since before you were born nineteen eighty five, according to our own Dan Curtis, and that's the Nasdaq
and A's at one hundred. What looks like it's not going to be a fourteenth day.
En it would be before I was born. There's a big focus right now on Google's TPU program. There's a report from the Information that there are talks between Google and Marvel about an inferenced specific TPU somewhere in the future. By the way, that's a report and story that Google is declining to common and full stop, but it's moving markets in the sense that Marvell is getting a big
boost from it. Separate to that, Bloomberg's reporting that Google plans to announce its new generation of TPUs at an event in Las Vegas this week. Let's get the details on that with Bloomberg's AI Infrastructure reporter Dina Bass. So much focus on the TPU program, right, Google's own use of it and then the expansion of it to third party customers andthropics. One that you and I have reported on, right, What are they going to tell us in Las Vegas
this week? What is the future of the TPU program.
So what we're reporting is they're probably going to announce an inference chip, you know, for running AI models after they've been trained. Thus far, they've been doing training and inference in one chip. You know, we are expecting and reporting that they're probably going to announce something separate just
for inference. At Google chief scientist Jeff Dean told me in an interview, Look, you know the way inference demand is growing called it now becomes sensible to specialized chips, more for training and more for inference workloads, and that they're looking at a bunch of things. Their chip chief, i mean VIDAT declined to tell me specifically whether they're going to announce that this week, but you know said we'd be hearing more soon. And this continues sort of
a trend. You know, in Nvidia announced a fast inference chip from what they'd acquired from GROC you mentioned before the Cerebrase IPO that's also at this point really a lowly in seafast inference play.
Meanwhile, the play has been extraordinary with TPUs and the adoption by many would call even rivals Meta likely wanting in on the Google made chips. So how are they broadening and where do you think they're going to be getting supply from? More broadly, there's a lot of reporting in the market, for example, that maybe they turned to Marvel versus Broadcom.
I know you can't coment on that.
Directly, but how are they thinking about their own supply chain?
Look, I think for them, supply, excuse me, supply is a problem. I was talking to Google Deep my n CEO, Demesis Savias.
He mentioned that as well.
Look, you have Meta, which signed they told us, a multi billion, multi year deal to use TPUs. They're just getting their first big tranche of them. They're trying to figure out what they're going to do with them. Anthropic has a huge deal, Citadel is going to talk about how they're using TPUs at the Google Next conference this week. And what Demis was telling me was that, uh, you know, contrary to what Jensen Wank said last week on the Dark Hush Fatal podcast that you know, it's really just
anthropic that wants them. They actually you know, have a lot of people that are interested. They don't have enough supply. Uh he was Demis was saying to me, Look, you know, what they end up doing is prioritizing the you know, top of the line frontier lab customers, because those are the customers who are most capable of taking advantage of what TPU has to offer.
You know.
More broadly, I think the TPU play, and this is why it appeals to these big frontier labs, is that Google is the only, you know, maker of a large, top of the line frontier model, one of the top you know models that also makes uh you know, AI accelerator chips in a large volume, has said they will as well.
I think that's why it's worth lingering for a minute on why the TPU is useful. So you explained really well that to this point, the TPU tensor processing unit has basically been a general purpose accelerator training or inference. But when you put it side by side, as many people try to do against Nvidia's latest GPU or other inference specific chips, you know what is it about Google owning the architecture, about designing it for specifically the inference
use case that makes it better. Is it money, is it power? What do we need to know?
Well, Google's argument is that it is the fact that they know what you need for training and running a top of the line model. There are you know, two things in the last couple of months that have really increased the interest in Google TPU. One is the Anthropic deal, which was a validation of the technology. The other was sort of the release of the latest version of Gemini, which was trained and is running its inference on Google TPU,
and the you know, strong reviews that it's gotten. Google uses data and requests and information from their own AI model teams to figure out what they need to prioritize and frankly, what they need to.
Fix in the chip business.
They you know, they work together to figure out that, for example, Google TPU wasn't doing you know, utilization was too low on those chips when you were using it for reinforcement learning. Demis was telling me they're using it to figure out, you know, how precise the chips have to be as opposed to where they can save money. And that's sort of a set of data that flows into the Google TPU design team that other you know, other chip makers don't necessarily have. And Vidia does have
a very you know, solid model team. But you know, among the big three anthropic open AI, Google, you know, for frontier models, Google is the only one making AI accelerator chips at volume right now.
Databas is a great story.
Can't wait to see what comes out at the Vegas event.
We appreciate you joining today. Now let's just dwell a little bit more on tech stocks.
They are taking a pause like the rest of the market today from their recent winning streak, and that's as investors navigate concerns so peace talks between the US and Iran. But let's discuss the tech outlook in particular with Anna Rathman, CEO of Grenadilla Advisory, and once again geopolitics comes to the forefront. But the fundamentals of tech are they vindicating the valuations because boy, if we had a run up of.
Late yes, I mean, I think it's difficult to value tech at this point, especially with AI, just because just a few weeks ago, pre war, we were worried about the six hundred and fifty billion dollar AI infrastructure spent by the hyperscalers, and today we're not really talking about that.
In fact, we're wondering if we have enough capacity. And you know, frankly, the previous reporting about Google's TPUs, I mean, we can have all of these deals right looking forward to twenty twenty seven, twenty twenty eight.
At the end of the.
Day, it really is about shortages, right, And this is where the geopolitics really comes in as well well, as you know, we're hearing it from the manufacturers and the ism manufacturing in disease. So I think that the valuation is really difficult for tech because computing computing needs are real, but you need to you need real assets in order to build real assets, which are the picks and shovels of the AI story. So I mean, I think we're kind of in a quandary here.
So does that mean on don't do anything s done pat until we get clarity at least on the geopolitical side of the equation, which in and of itself affects the supply chain of chip manufacturing. For example, helium that can't get through will mose right now.
Yeah, I think AI is one of those stories when it takes off. It takes off, and it's going to be difficult to catch. So in terms of investing, I think you stay invested. If you are invested in hyperscalers, you're already exposed to a lot of data centers. Perhaps think of some of the adjacent industries of manufacturing and industrials. As far as the war is concerned, it's difficult to
price that in as well. I mean that the story is very different today from last week, but I do think that the markets have sort of priced in not a binary outcome of will we have a deal versus will we not have a deal. I think what the market is pricing in is this very wobbly, very messy ceasefire that might continue.
For a while.
So I find this really interesting. We've spent a lot of time emphasizing that then as that one hundred had this amazing thirteen days streak, and people seem to forget that before that thirteen day period there was a big sell off in technology shares, and if you look at like some of the Goldman Prime brokerage data, you know, hedge funds were dumping out of US tech stocks software
in particular. Why is it do you think that the investors sort of then felt that the calm to start buying almost immediately as the war was taking place.
Well, you know, I don't think the investors were waiting for the calm. I think the investors were waiting for some bottom where it felt cheap enough to come in, because again, the AI story is a long term story.
Even software.
I think it was more of a sledgehammer tool that we were using to knock out all of the software indiscriminately. And now software is still going to be pressured in my opinion because of AI AI native AI stories, but I think at this point we may be applying more of a scalpel rather than a sledgehammer.
So from that.
Perspective, everything was oversald. So I think investors just look for a cheaper point to enter the market.
Technology earnings start big time in Earnest this week in some sense. What do you think will happen? What do you think what kind of bar do you think is set for investors for this tech earning season.
Well, you know, if we look at some of the international tech stocks that we've heard from, such as TSMC and ASML, I mean that demanded story.
Is here and real, and you know in the future as well.
So I do expect a demand story to continue. Whether or not the companies are actually poised to capture that is another story. That's a corporate story, that's a management story. So one in particular that I might be a little bit nervous about would be Intel just because of the recent history, just because of some of the issues with the foundries. Other than that, I'm still pretty constructive on Tech and we'll see what they actually report this week.
Anna Rathbun and Grenadilla Advisory, thank you very much. Now coming up, Blue Origins New Glens successfully reuses a booster for the first time, but the mission also delivers a major setback for the payload that was on board. We're going to hear from Blue Origin CEO Dave Limb Next, this is Blueberg Tech.
It was a.
Story of both success and failure for Blue Origin Space launch over the weekend. For the first time, the Jeff Bezos company reused a booster for its New Glen rocket and landed it successfully, but the satellite it was carrying for ast Space Mobile wasn't placed in the correct orbit and is quote too low to sustain operations, and.
It was painful if you own the shares of AST Space Mobile. Therefore, look rocketing back down to Earth down nine percent for AST's Space Mobile.
At one point it.
Was off by fourteen percent. And this off nominal orbit means a lot for a company that needs to get about fourteen up in this year.
Yes, you know he is a paying customer of Blue Origin. And if you go on social media and you look at Jeff Bezos or Dave Limp, the CEO, they were celebrating the Blue Origin part that they landed the booster. They reused the boosters only the third time New Glens flown.
But basically the satellite that they were sending into orbit was in the wrong place and it doesn't have the means in terms of frousters to move itself, so they deorbit it and the insurance company actually comes in and pays for it.
And that's what's word about AST Space Mobile sort of being the one that bears the brunt here because they get the money back from the insurance, but it needs to get its satellites back up there to be able in a swift succession to allow you and I to
make the calls that we want to do. Look Scotiabank analyst Dress Concelo was sort of saying, look, even though the reusable Blustellan is successfully, the mission was a failure, and the Bluebirds seven failure highlights the need for well more redundancy backup plans right for these failing units.
For it's a black eye for Blue Origin because they part of their business plan is to put satellites into orbit for customers. But again, they really needed the rocket itself to work, and it did. I spoke to Blue Origin CEO Dave Limb, but that was ahead of the launch. We talked big picture about the company's future, and this is what he had to say about the new GM program.
I think it'd be a really good year if we could do you know, eight to twelve flights this year, up from two last year, that would be that, you know, that'd be a good cadence for us. I can see a path that we have plenty of hardware to do that. It really is what we learn from the flights.
I would tell you that there's there's never been.
A time where launch is in such demand, and by the way, I think the demand's going up with all these directed device announcements Amazon made one obviously, other AST is doing directed device and these mega constellations. We have our own constellation. We announced with Taro Wave that laun if I if I could fly a rocket every week right now, I would be sold out for the indefinite future. And so our job is to get out there, get as fast a cadence as we can, as quickly as possible.
And the foundation of that, again is just to build a world class manufacturing business. And you know, that's something I've done for a long time, and the team's done for a long time, and I think we're on the right path. We still have ways to go, but I'm I'm excited about the progress.
The world became familiar with Blue Origin through New Shepherd, the smaller form factor launch system, and the first focus of that was kind of space tourism, you know, going to a pretty low altitude relatively speaking, for a few minutes of weightlessness. And you know, of course I was in Van Horn, Texas when Jeff Bezos himself went up. But you took the decision to halt that program in large part to focus on Blues participation in getting America
back to the Moon. And so you know, in the first instance, Dave, what has the result of that decision been. Have you been able to accelerate that other program, the lunar program? How have you moved and reassigned resource and people to fulfill that ambition?
Yeah, you know, we were able to fly ninety eight people above the Carmen line. I've never seen so many smiles on every astronaut we were able to do that for And I think if you had to make the decision to put New Shepherd on pause, we put on pause for at least two years. With your heart, it would have been a very hard decision because you see that, you saw it when you saw Jeff fly. Every customer
is the happiest customer you've ever seen. But when you make it, when you make that decision with your head, it was probably one of the simpler business decisions that I and Jeff have ever made, because you know, I am so passionate about.
Getting the US back to the Moon, as is Jeff.
It's you know, I don't think the country wants another spot Nik moment. We have been saying it blue for twenty plus years that the Moon is this incredible gift given to us. You know, it's three days away, It's got all the resources that we'll ever need right there at least for the foreseeable millennium ahead of us, and.
We know how to get there.
Through Origin CEO Dave Lim there speaking with it over the course of the weekend. And coming up we are talking Apple's revab criy.
It may have been revealed more on that. Next there's a broombag tech.
Time now for talking tech and first up, China's IG says it expects AI to produce the bulk of its film and TV content within five years. Now, the Netflix style streaming platform plans to transform its app and website into a social media destination centered on AI generated programming. Now, as part of that shift, the company unveiled a new AI talkit it says can handle nearly every stage of
the filmmaking process. US regulators across Asia are taking a closer look at cybersecurity risks in their financial systems as concerns grow over the spread of appropics mythos model. Now in Singapore, the financial regulator is urging banks to tighten up vulnerabilities, while in South Korea, government agencies have been meeting to plan their response, and Huawei has unveiled its latest folded phone, the Pure xmax.
Now the device.
Unfolds horizontally to rough it the size of a small tablet. There's priced around oney six hundred and ten dollars now.
The launch comes as.
Competition with Apple heats up, but Apple expected to roll out it's own affordable later this year.
ED Let's stick with Apple, whose upcoming WWDC event is set to reveal the future of Siri. But ahead of the forum, Apple's teaser may have revealed a glimpse of the revamped interface. Let's get to Bloomber's Mark German, who covers all things Apple and consumer tech, and this was the subject of this weekend's power On. I don't Mark, I don't know. Every weekend I repower on. It's usually
based on who you're talking to, your reporting. This time, you're getting your your your microscope out, you're looking at the imagery. What are you trying to learn here?
Well, you know, every time Apple announces an event toor they announce in your conference, they put out this teaser, graphic or invitation, and from time to time it gives you a pretty good indicator of what's to come. But usually you only realize what that actually was in hindsight. But now I can tell you, you know, several weeks ahead of the conference that this WWDC font, this logo, this glow, this design they have there actually is representative we're.
Looking at it now interface.
Yeah, it's a representative of the new Siri interface. Basically, they're moving Siri to a new design. If you're familiar with the iPhone Dynamic Island, when you trigger Siri, it'll peer up in that little camera area and then it has that new font like you have a WWDC logo and it has a glowing cursor similar to the glow you see around the twenty six and that logo you were showing.
So fold that into what we experience, how that will seem what in fact Siri will seem like when we finally get it into our hands or our ears or our voice, the way we're going to interact with it.
You know, Siri has looked cool for for for many years, but it hasn't worked so cool for for many years.
Right.
It's it's in many ways, it's just pushing people to the design in order to distract from the functionality. That's what we saw from Apple Intelligence. But this time around,
they think that they have nailed it. They think they're going to get Siri to work with the features they previously announced, some of the new enhancements that are coming that they haven't announced yet, or a Siri app so a standalone app, a chat bot like interface for the first time to better compete with Gemini and CHACKGPT, and then a new feature that they call internally World's Knowledge Answers, which is basically a perplexity competitor, so you can search
the web and get summaries of information without it actually pointing you to Google. It'll happen within the Siri interface itself. So if this all works, I think they have a pretty compelling story to tell for their voice assistant in AI.
Come jack Mark real quick. It's better to follow this story and aggregate over a number of years, Like I think back to two or three wwdc's ago and you reporting like Siri and Apple Intelligence isn't going to be the focus, then looks like it will be the focus. Now summarize where Apple was actually landed with Siri, that is AI improved.
Now In terms of the underpinnings of Siri, they're using technology from a Gemini model, merging it with their own model technology in order to get it work more properly, it'll be able to analyze data on your phone to help you take action on it. But big picture, this year's conference is really all about Siri and AI. They basically hit pause on all non AI or serre related features. Lots of bug fixes and performance enhancements, but the focus is on this voice assistant this year.
Mark German is always giving us the hints. We so appreciate it.
Let's talk about now humanoid robots for a minute, because one has just won a half marathon race for robots in Beijing.
But get this, it ran faster than the human world record.
The race was held in Beijing over the weekend, showcasing China's really rapid advances in robotics and artificial and intelligen and the robots strode across the finish line in fifteen minutes and twenty six seconds to the state run China Daily. That's according to them, and that beat the human world record holder Jacob kip Plumo of Uganda, who finished the half month and in fifty seven minutes and twenty seconds last month in Lisbon. What a stunt, I mean Last year it took more than two hours.
For the robots to win.
That's the video of robots running in a race. I mean that can we let's just literally what is one hundred teams? Three hundred robots? The director won't show it, but I'm just grinning and laughing because that's a video of robots running in a race.
Karen, what was important?
Liquid cooling that's going to go on in that Maybe that helps us in other parts of window.
So many DPUs on fire?
Okay, bloomg Or, welcome back to Bloomberg Tech.
We talk about geopolitics. Now.
President Trump says it's quote highly unlikely and a ceasefire whether RON will be extended.
This is VPJD.
Vans is set to head to Pakistan for deal discussions. Now, Bloomberg's Jeff Mason has just spoken with the president. But a few moments ago, you join us now, Jeff, and this tone seems negative.
You know, I'm not going to say negative or positive. Just what he said, which was a couple of different things. He said that he thought talks could work out well for all parties. But he also said, as you rightly quoted, that he was not intending or it was at least highly unlikely that he would extend a ceasefire. He also said that the ceasefire ends on Wednesday, which is a little bit of news because the two week period that most people were assuming goes through tomorrow Tuesday, so that
was interesting. The President also made a point in our phone conversation of saying that he was not in a rush to get a deal, and he talked about the length of the Iran War. The excuse me, he talked about the length of the Vietnam War, the Afghanistan War, the Korean War. So he was watching this timeline or his own timeline, in the fact that many other wars have gone for a long time, and he just said he was not in a rush.
I'm just going over the headlines that were generated from the interview, Jeff. There are several of them. I guess operationally, what happens next? What did the president tell you about what he and his cabinet will do in the coming days. We mentioned jd Vance, the Vice president, and what he does now plan to do with regards to talks.
Well, he gave a little bit of detail about when the talks are expected to happen. He said that they were likely to be on Tuesday night and Wednesday morning, and he said that Jade Vance was expected to leave later on in the day today Monday, which contradicted what he said to another reporter at another news outlet earlier, where he suggested that the Vice president was already en route.
So the next step are these talks.
I asked him when we spoke about the ceasefire if he expected that fighting would resume right after the ceasefire concluded if there wasn't a deal, and he said he expected so. So to your broad question, it's kind of a dual track here. He's expecting talks to take place, and he showed some optimism about those talks, but he's also clearly holding over the threat of additional attacks to start as soon as this week if a deal is not concluded.
And just for a reality check.
For our listeners and viewers, remember that the last time there was a major piece deal with Iran happened during the President Barack Obama's administration.
That deal took a year and a half to hammer out.
So timing here is it would be unusual if they're able to get a deal as fast as Wednesday evening, which is what the President said was the deadline for the end of the Seaspire.
Bloomberg Jeff Mason, who's just off the phone with US President Donald Trump, Thank you very much. There's a lot going on in markets, much of it relates to the headlines about the war in the rom but actually in bitcoin's case, so we're around seventy five, six hundred US
dollars partoken Carrot. I saw the headline that Strategy Michael Sailor et cetera brought two point five billion dollars in bitcoin in the trailing seven days, and I think that that's their biggest buy since the end of twenty twenty four something like that. Maybe that's a factor in the market. Maybe it's just a risk asset behaving like everything else.
Who knew that Michael Sailor actually still wants to continue to buy bitcoin. I mean, many wouldn't not be surprised by that. It's interesting that in the defile space more broadly, that's been that we can hack training what was three hundred million.
Dollars, which has helped those consequences. Right, It really.
Does just negative sentiment around DeFi. But people can separate the two. Your bitcoin is not DeFi. DeFi is not Bitcoin.
Really quick in the technology space on single names, Adobe is higher by about one percent. It announced basically a broader agentic ecosystem with basically everyone in the world of tech aws and thropic, Google Cloud, IBM, Microsoft and video and open Ai, and an earlier report on that had boosted shares. And then there's the kind of big thing that's happening in aggregate. The Nasdaq one hundred is softer and actually push to session lows after some of the
headlines from the President. But the main thing is it snapped a thirteen day winning streak, right, Karen, That is what we've been talking about in that thirteen day period, sort of how tech is unstoppable in the face of what is sort of a tense situation in the Gulf that speaks of Benberg equities for water Carmen Rhini Key, I mean, that's kind of the story, right, Tech has cloud on irrespective of those geopolitical overhangs.
Yeah, we've seen just an incredible rush back into risk assets from investors. I mean even today, you know, we had this really significant geopolitical news, at least at first you said Drum wasn't quite as significant as I might have expected. And there are parts of tech that are actually still holding up today. So I was writing about the IGV the software ETF last week it had its best week since two thousand and one, really significant gains there.
Today it's actually holding up a little bit. It's doing a.
Little bit better than the hardware sector than the socks. At least last that I checked on my terminal, it was up a little bit today. Hadn't completely erased doated games exactly exactly, and so we are seeing still some bright spots here. And I think the names that drove the nasdaqs thirteen day streak are really interesting. So we definitely had talked about Intel had such an incredible rally that.
Was the top of the charts there.
It was the number one performer over those thirteen days, and then the memory names were right after it. So Marvel, sand Disk, Western Digital.
Can we go into Marvel a little bit more because the hardware story is basically.
What's been unstoppable.
Yeah, and Marvel that seems to be reporting out there that maybe they're going to be working alongside Google, Google broadening out from its TPU broadcom relationship to some of the players.
Yeah.
Well, and this is one that actually the stock is sort of bucking the trend today, at least in the inter day, was on track to close at a record high.
So another record high. Marvel's still looking really strong.
And yes, so the news today is that there are reports that it's in discussions to develop two chips with Google. So one is immense memory processing chip and the other is a TPU.
And those are both really significant.
It's also potentially intox aimed at designing semiconductors exclusively for Google. So that's a huge deal for a company like Marvel. And again, you know, it's had a really incredible rise. So through Friday's close it was up more than sixty percent this year. It's just extending those days.
That game all calms down in the session as well. Right, you both got to point out it had an astonishing start to April, like every other member of that basket. So interesting move.
Yeah, we've got to keep an eye on the hardware and we can dig in a little bit more.
But common Hinikey.
Sets it up perfectly, always got her eye on the broader moves. Let's dig in a little bit more on this Marvel's story. But in my intelligence, senior analyst Con Savanni, who has been watching the stock positioning itself for maybe taking advantage of custom chip demand. And here we have it, maybe some reporting showing that they can be really relied upon by the likes of Alphabet's Google chip design. What do you make of the reporting coming out of the information in this particular instance.
Yeah, this reporting really validates what we have been talking about for the past two quarters that you know, the noise prior to six months in Marvel was all about Amazon Tradium three. But we think they're really expanding beyond just the number one, number two hyperscalers. And what's more importantly is here is that MPU which just falls in the x few attached portfolio. This is the areas where sophisticated hyperscalers like Google, Amazon, Microsoft are now optimizing customs
chips beyond just the core powerful TPUs or accelerator. So this is where we think Marvel can gain significant shell and this comes at higher growth margins.
By the way, Yeah, in inference memory is more of a bottleneck, right, and so the MPU just moves data, you know, that's its purpose. I think the thing that we try and reiterate is that Google owns the design and architecture of TPU either way. So in that case, what is it that broad Com to this point and potentially Marvel actually do conjem. What's the value add of partnering with someone like that.
So when you think about beyond that core digital logic, it's the IDP sets surrounding. How do you connect to the memory, how do you do your packaging, how do you have your intercorrect iOS. This is where these companies come in with their own IP portfolio and add value. One more thing to add here is the influence hardware streck is changing. It's not just one hardware stack fits all sides, so we are going to see multiple players.
There's also talks about media tech also participating in one of the Google's sort of keep you versions, So we are going to see this diversification when it comes to ASIK now.
And by the way, on the Marvel side of this, Google declines to comment to Bloomberg on any of the reporting. But it's not Google, it's Broadcom or Marvel that phones TSMC and gets those things. Fat Congensabannie from Bloomberg Intelligence with the research. Really appreciate it. Now coming up the conversation with Greg Martin of Rainmaker Securities, as Markets get set for some mega IPOs. This is Blomberg Tech. AI
shipmaker Cerebra Systems is back in the IPO spotlight. The company filed publicly for a US IPO on Friday, just six months after pulling its previous attempt, so to say. The deal could raise around two billion dollars, testing investor appetite for large scale listings. The company's filing shows gross margins slipped to thirty nine percent twenty twenty five from forty two percent a year earlier, highlighting some of the pressures. Even high profile AI named a facing character at LISTA.
It's not loss making this year for versus last year. Let's well, revenue certainly doubled.
Let's talk a little bit more about the IPO landscape and joined with Greg Martin, Managing Partner and Maker Securities, and Greg, what is the appetite likely to be?
Say we go into Cerebraus and.
We think about how, yes, they've suddenly galvanized revenue growth, but the margin has been hit.
As Ed articulates, are people just going to be.
Betting still on this desire for AI? Compute the rampant bottlenecks that we see in terms of just access for AI iterations and innovations.
Well, it's a great question, and clearly this is going to lead off, you know, what could be a big AI infrastructure year. There's already some public companies like core Weave and Nebius who had nice run ups recently, so there's clearly demand for AI infrastructure and Cerebrius numbers. They've they've grown revenues. I think the biggest question mark with Cerebris was customer concentration. Now they've added open Ai, but they still have customer concentration, and so it's also a
very high valuation. The company did five hundred ten million of revenue last year. Their last round was at twenty three billion, which was just done in February, and so it's meantime, the secondary market for Cerebris is actually traded down to about ten billion recently, and so there's a big question mark of whether they can hit a big enough number in the IPO pricing such that their last round investors won't have a down round. So I think
the market we're watching Cerebris closely. It's not the marquee name like SpaceX and propic or open Ai, but it'll be interesting to see if the market really is open again for a company like Cerebris with such customer concentration.
And this is why we go to greg because we think about the quick that you can see, you can tell our audience about in the secondary market, just how hot is it for the really exciting names like we were just reporting Edge and Team about how Anthropic basically is turning away more money in terms of a capital raise that could drive its valuation up to eight hundred billion.
Also, yeah, and Thropics one of the hottest companies we've ever seen. Unbelievable demand on our platform or Rainmaker. Obviously, the company added ten billion of ARR in one month. They're now up to about thirty billion of arrs reported publicly. So it's just it's people are starting to value Anthropic as a share of the economy, not in sort of traditional metrics that you might see. And of course their growth actually does start to rationalize the valuation that they
that they've had. We're starting to see them, you know, enter into just about every software category. So Anthropic is about as hot as companies we've seen. They raise money at three hundred and fifty billion just in February. We're now seeing potential eight hundred and we're seeing unlimited demand on our on our platform. So clearly Anthropic is the hottest company we've seen in a long time.
You know, with those names SpaceX, Open AI, Andthropic, people always ask when we report something, you know, how do I, as an everyday person invest into those companies? And the first thing I say is, we don't do that. I don't do financial advice. Just read my reporting. But the second thing is I say, oh, but we have reported that SpaceX, Open AI, and Anthropic are all likely to go public in twenty twenty six. And with that flood into the market of three names at that scale, how
does that translate to activity on the secondary market? Do people try and position themselves in anticipation of a big listing or public debut moment like that?
Yeah?
Absolutely, you know, these three companies in particular, there's they're obviously you know, enormous market caps eight hundred billion, two trillion potentially for SpaceX. So we're seeing a ton of demand pre IPO because people do think that there's going to be a run up in the markets. All of these companies are oversubscribed in the secondary markets, and as companies get closer to IPO, there's a feeling that hey, now I'm going to have liquidity, Now I'm going to
have access to more information. But we've seen in the past IPO allocations get cut back substantially, and so if you want to get into these companies with any kind of scale, you may have to come into the private markets and buy it via secondary or primary. So we're seeing rabid demand for all three of these companies in the secondary market.
Greg really quick. We had lots of little IPOs last week, many of them came on this show. What did that signal to you?
Not a lot, you know, I'm much more focused on the bigger IPOs. There's clearly a massive backlog and so companies, there are companies that just need to go public. I don't think there's a lot to be you know, learned from what we saw last week. Listen, we have the potential for one of the biggest IPO years in history. The three companies you mentioned are going to potentially raise up up close to two hundred billion in between them, which could suck the entire oxygen out of the market
for a while. So I think what's really going to be interesting is how those three companies, SpaceX and profit and open AI perform, And of course it's a function of whether the ip OF market's truly opened, because I don't think the market's open until the streets and corms are open.
God, if that's what you're banking and waiting on, we'll be here a while. Greg Martin of Rainmaker Securities secondary market broker who sees what's happening in terms of liquidity interests ahead of IPO now coming up. Nick Stronsky, co founder and CEO of Revolute, discusses the digital banks future plans on this week's David Rubinstein Show. Nick Ronsky built Revolute into one of the world's most valuable fintech companies,
now worth around seventy five billion dollars. The co founder and CEO speaks to David Rubinstein about the company's future.
I think it's obviously much easier for us and giving on your administration. Plus we have so many other Manking licenses. Plus we have a Manky license you can now, So for us it became much easier compared to two years ago.
So in the United States, what do you hope to be able to do with your banking license if.
You get it?
What is the market gap that you think you can feel that maybe others aren't doing.
So basically I mentioned a radioh retail business. So it's one simple up where you can do everything what you need for your financial life. And then second direction, we'll also provide the regular business accounts for businesses. And again it's end to end services that you need to run your business. So payments, banking, credit will come as well, wealth management or treasury we call it treasury, and then
acquiring as well. So for example, right now, if I'm an American company, I'll need to stop say Strip for acquiring, then Gip Morgan for my business banking. Then you know, something else for payouts, something else for payroll. So we provide all these services in one product.
Well, if you conquer the US the way you've conquered the UK and Europe, is China next, or you're not going to focus on that.
Easy way that if you're focusing so spending a lot of time or studying Asia and then as a search, so we apply it for multiple banking licenses in Asia as well, and then some of them but we got already and then China is on the list as well. You know, one day I hope we can set up a business there.
Now some people have written that your bank Revolute is worth seventy five billion dollars. Have you ever thought of taking your organization public?
But the way I think about it, will be ready to probably into year's time. But then again depends on how good the market is.
Bench you don't have an IPO, how do people who work for you make some money? You make a market for people to sell their shares internally.
Yes, so we do secondaries every one in two years when people can sell certain portion of their stocks to accel investors.
So you've done secondary sales for your employees, so employees can then some liquidity. Do you expect before you go public you'll do any more of those? Or you're just going to wait till you go public before you liquify any more our stock.
We all think, so we might do second risk as well before we do ip your because we do it every one or two years, so we might do a second result.
All right, Well, if anybody wants to sell their shares to me, let me know. And have any investment bankers told you why they're the best to do the IPO for you?
Yeah, obviously every single one is the best.
Everyone is the best. You're not in a rush to do it.
You don't need the money, so we're bank.
And then for the bank, it's super important to have trust. And then public companies are trusted and more compared to private companies.
Go see the full interview with the Revolute co founder or CEO, Nick Sterronsky. Watch the David Rubinstein Show to peer Conversations on Wednesday, April twenty second at nine pm in New York and you O'm been about television Now more on digital banking JP Morgan City Group. I've taken their rivalry onto the blockchain, competing to define the next roun tier of global payments. Just take a listen to what they told us in a recent joint interview.
The real client problem that we were looking to solve is the ability for large multinationals, big banks and broker dealers and FinTechs to be able to move their money around make payments seamlessly around the world.
Twenty four cent.
We want to keep building on the public chain stuff.
We have now tokenized money market funds with our asset management business through Conneccess. We have tokenized deposit tokens, we have a full tokenization engine we have core infrastructure, so we basically want to want to battle hard net.
Rivals on screen, and the two banks of building competing systems but takingly different paths. Blue Megs as wellly is here with one of the most red stories on the terminal today as well.
How are they doing it differently?
One like stable coins, one doesn't so much exactly.
So it's interesting to see Wall Street raised into this space, and it's even more interesting to see their approach. So for a city, they've signaled an openness to stable coins. They've partnered with coinbase to really push into this digital asset space and they want tokenized securities. Around five hundred clients have been enabled in that platform and they process around one billion dollars already.
The JP Morgan day, isn't it?
Yes?
JP Morgan has been a little bit more reticent when it comes to stable coins because for them, the KYC is a little bit less. What they prefer is to build their in house infrastructure. Their in house blockchain called kinnexis and Cannexus is really already processing over three trillion dollars in volume, with each average daily block being five billion dollars. So to be sure, this isn't really grand.
Yet in the larger scheme of things, they process trillions and trillions in their traditional payment system, but it's growing.
So they're preparing for a world in which everything is tokenized. Right, So we're playing up the kind of frenemy I suppose, angle of it. But what they are doing is laying groundwork for the state of the financial system that doesn't quite exist yet.
That doesn't quite exist yet.
We have the IMF, I think two weeks ago, warning that if we build too much into this space, we might see actually a great crash because in theory everything works send a token from two AM and Maldives to your friend in New York. But in practice it's still a little tough, and especially if a lot of these from the Nasdaq to the New York Stock Is Chain, to JP Morgan to City build their own worlds, it's going to be even more fragmented than what we see today.
So lots two still parse out in this space, but it's exciting to see this big big banks really push into token is NAFE token ie real world's assets, real world Yes.
Bloombergs Isabelle Lee with the story that JP Morgan and City squaring off on the Next Payments Frontier. So that does it for this edition of Bloomberg Tech.
Yeah, don't forget to check out our podcast.
Why not.
You can find it on the terminal, if you're on a line on Apple and Spotify, on iHeart ed.
Is here in New York Nantah week Right, Well, lucky.
Lady, a fifty six minutes into the show, if you've just worked that out, congratulations, it.
Might be a holograss. It's a push into.
That recap on the pod, and it's nice to be here with you in New York City. This is Bloomberg Tech.