Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is alive from coast to coast with Caroline Hyde in New York and Vla low in sent Frances Go.
This is Bloomberg Tech coming up. AMD shares sync after a sales forecast that underwhelms investors, the chip maker perhaps not making the aim roads some on Wall Street had hoped for.
Plus, the sell off continues across tech markets as fears of AI disruption they take hold.
Following Anthropics new automation model.
And Bloomberg reports Nvidia is nearing a deal to invest twenty billion dollars in Open AI, with the AI company looking to raise up to one hundred billion dollars in its new funding round. AMD is ugly right now, the stockdown fifteen percent. It is on track for its biggest drop since October of twenty eight. That is how severe this selloff is. Let's get to the numbers. It's projecting sales in the period roughly nine point eight billion dollars
plus or minus three hundred million dollars. That was above consensus estimates. But go into your Bloomberg terminal and you'll see at the high end of the ranger's forecast. Many expected sales to be ten billion dollars. Anyway, then there's the China issue. They had China revenues, but it weighed on Margins AMD's results. Let's react, Congensubanni, a Bloomberg Intelligence who in his React piece in his research you talk about the impact of China. Good. We recognize some revenues.
It impacted Margins. You talk about the CPUGPU ramp that's not coming until the second half of the year. Take it from that.
Yeah, I mean, fundamentally there was nothing wrong in this report. Is just when you take out the China revenue was a three ninety million and one hundred million for the guide. The beat magnitude was really modest, and that didn't hit at part of the lofty expectations.
Again, most of this was known.
So very surprised to see a stock reaction that we're seeing today.
So the surprise what in terms of what enthusiasm had been baked in that shouldn't have been. How far had the market moved past already the bullish.
Tone that we get time and time again from Lisa Su.
Yeah, it's about it's about now waiting about the next GPU inflection point, which is which we knew is not going to happen in this quarter or next quarter.
It is for second half twenty six.
Remember, there has also been that noise in the market around it delays with their ramp of the four hundred and fifty series. None of this was shown in this report. I mean, we don't expect significant upside right now because most of their customers are really waiting to pick up that server revel Helios Raax.
Solution which dollar terms right, sequential decline in revenues maybe, and no one wants to hear that, right, And we'll wait until the rest of the year. Second half the year can mean June, it can mean December. Who knows a big piece of M and A in your industry? Texas Instruments buying Silicon Lab seven point five billion dollars. What do we need to know about this in the analog space?
Yeah, this is surprising. The last time they did a deal like this was in twenty eleven. Strategically this fits will they add a missing piece which was wireless connectivity in TI's portfolio. But more important the implications are financially. The key thing you'll remember here Texas Instrument states on
free cash deliverability. So this management has said will not impact that free cash flow, will not impact that dividend, and they'll be able to absorb the capacity in the capex that they've spent over the last three to five years.
Quinjinspani Bloomberg Intelligence and All Things Ship. We have breaking news on Sara Bras Cerebra Systems raising a one billion dollar Series H at a twenty three billion dollar valuation. This is a company that we've been tracking closely. Will they won't they IPO? Well, right now they're raising around a billion dollars at a premium valuation of twenty three billion dollars. We'll get more carr as we get that throughout this day and this week. There is plenty more news hit me with it.
There is because those fears around AI just they're continuing to ripple through the markets.
At the moment.
Ed Anthropic it accelerated the sell off in software stocks after it rolled out that new AI automation tool. Namba Gequity supporter Carmen Ryanikey has been.
Following this story. I mean the numbers are Goganshman.
I think it's somewhere like two trillion dollars been wiped off that Gold and SAX Software index more broadly since its highes.
But what can stop the rot right now? Because this has gone global? You know, that's such a great question.
And we're really seeing such a broad sort of reaction to the software stop. I mean, we know that software has been beaten down. We're seeing more you know, sort of stocks kind of join in the fray off of this latest anthropic tool. But it's also kind of spreading into the hardware stocks. We're seeing in video down, you know,
Google down. Interestingly, Microsoft is up a little bit today, sort of rebounding I think from a selloff from its earnings report, but it has spread even into the European markets and others as well, so it looks a little bit like a washout. I think I have heard investors start to say that maybe this is a place to start nibbling. You know, there could be really a rally
if software sells off too much. You know, many stocks in the index are in over sold territory, so we could see you know, some rebound, some reinvigoration in that trade. I think investors are really just sort of thinking, is it worth trying to catch a falling knife?
Colmin bring us some of the action of today Wednesday session, right because, like to be fair to our audience, this is a conversation we had yesterday. It is a multi day story that's playing out. What's new in the market this morning and what are we looking for next?
Yeah, I think that there are multiple things going on here. So we're not only seeing you know, this pressure on software, but we've had a couple of few earnings reports that have stumbled that I think are adding. So Microsoft up a little bit today, but that report, you know, injected some jitters into the market.
And also AMD as well.
You know, it really showed that even if you have a pretty solid guidance, it might just not be enough for investors to continue to buy your stocks. So we're seeing that as well. And you know, coming up we have Google report or Alphabet reporting tonight, we have Amazon tomorrow. So there are just a lot of catalysts that are coming into the market right now that are shaking investors confidence in the sector.
Of course, many pinpointed Anthropics legal offering, but we saw Google out with its gaming offering that had seen a lot of hit to some of the gaming software names. But I'm also interested in we're a public market identifying show right now, but private markets, in particular the debt markets, we're seeing them roiled as well, and it's impacting certain pea houses exactly.
Yeah, So there's just been a major softfare as well. This, I mean, this idea of contagion I think is huge, right and you know, the pain here from software spreading into all these other parts, you know, not just equities,
is really prescient right now. So until there's sort of this idea that there can be a rebound, I think it's it's likely that we see that continue to spread and investors sort of continue to question what's next for the sector, especially if AI is going to eat the lunch of these companies.
And by the way, guys brace earnings is still happening, right, So, like there's all this news flow and there's market moves on on a sort of macro concern about AI labs eating the lunch of software companies. Alphabet's going to come out and it's going to tell us about how it's doing. I mean, how does that function into your week? Fit into your week? Sorry, calm and the earning story.
I mean, they're going to be huge. I think the thing that we're looking at the most for Alphabet and Amazon are these cloud numbers. And that's going to be really important because that's where we saw Microsoft stumble. So that's what I'm going to be really laser focused on
to see. I mean, obviously all of these businesses are different, there are other components at play with both Alphabet and Microsoft or sorry hands Amazon, but you know the cloud figure is going to be absolutely the most important thing and could really move these stocks in either direction going forward.
Bluomo's come rhinicky all over the software space. Thank you very much. Earnings again, shares of Uber. This is so interesting. We're down three percent now in the pre market. We kind of swung to a gain. We've been down like nine percent in the pre market. I'd call this a mixed bag. The company posted a weak profit outlook and announced a new CFO with a focus on a robotaxi future and new markets for robotaxi. Bluebos Natalie Lung covers
consumer apps, theog economy, and increasing the robotaxi. Where do you want to start? Like, I think, let's go with the financials, right, the earnings. They gave this a justed EPs forecast. Ebit Dar forecast that appeared light, but actually management that explained some of that already just explain.
Yes, so the EPs outlook for first quarter came in a bit light. That could be a result of a lot of their new product initiatives, let's say the low costs shuttle product at airports and even the high end you know, premium reserved rise at airports. Sort of these new products not yet yeuring the return that they're hoping to see, but they're still hoping that would sort of
expand towards this year. That said, like the fund mentors, the demand underlying demand, especially in the US where they have some of their most profitable markets, is very healthy. They think grow through accelerate there as some of the risks we saw last year, like high insurance costs, they think that will sort of abate this year.
And growth bookings can be up to fifty three and a half billion dollars for the current fiscal. Call to Natalie Donokles with Shahi mainly trying to talk up so the healthier pricing, what about this new CFO, what about the focus on robotaxi from him.
It's interesting they appointed that they promoted Pilaji Christiaan murmurthy to the role of CFO. He's one of the more outspoken, tartomous bos among the leadership. He's always on X defending Uber's AV strategy, and so we'll definitely see Uber double down on that sort of narrative, you know, trying to combat what it caused misconceptions among Wall Street bears on how it's got to you know, facilitate and support the ecosystem rather than completely be disrupted by rivals like Tesla or Weimo.
And where Mo they partner with often as well, Natalie Lung, We appreciate the breakdown and talking of Avis. The Senate Committee on Commerce, Science, and Transportation is holding a hearing on the future of self driving cars right now to assess existing autonomous vehicle regulations. Of course, this is as the race with China to lead the future transportation technology heats up. Witnesses are from Tesla, from Weimo and others
they're participating. If you want to follow ono, just go to Live go on the Bloomberg terminal D.
That's Tesla's Lars Maravi on the on the screen right there will track the headlines coming up. Jessica Lujim from the Competitive Enterprise Institute joins us to talk about antitrust concerns from a Netflix Warner Brothers tie up. We have more on that in a moment. AMD sell off accelerated in the last few minutes at one point down sixteen percent, already on track for its biggest decline since October twenty eighteen.
A forecast for sales in the current period that at the high end of estimates was disapp pointing, but also a sequential drop in revenues. And we are waiting on this big GPU and server design moment that isn't coming until the second half of this year. This is one we're going to watch for the hop This is Bloomberg Tech.
This deal keeps one of the most iconic Hollywood studios healthy and competitive. Warner and Netflix together will create value for consumers, more opportunities for the creative community, and more American jobs.
Netflix co CEO too surrounders there testifying with four lawmakers about the company's plan to buy Warner Brothers streaming and studio divisions. Surandos also said yesterday that the deal would give streaming consumers more for less, noting the eighty percent of Netflix subscribers also pay for HBO Max currently owned by Warner Brothers. Let's talk through the anti trust concerns with Jessica Manujin, director of the Central of Technology and Innovation at the.
Competitive Enterprise Institute.
Jessica, you've actual, we've testified before Congress about streaming competition in previous times. So what did you make of Ted Sarandels and his argument that this is going to give more consumers more to them for less.
You know, I think you can buy it. There's so much overlap in people who maintain a Netflix subscription and also in HBO subscription. The companies have talked about how they would be willing to combine those and offer those at a lower price. I think he said as much in the hearing yesterday. So there are competitive concerns that remain for some people, but I think certainly there are efficiencies in overlap. Because the most of this merger is
a vertical merger. There's pieces that one company doesn't have that the other one does. They're going to fit them together and that could be substantial cost savings that one hopes get passed along to consumers.
What was interesting amid the politicking that goes on on Capitol Hill in these sorts of hearings the time and time again, iteration we hear from our guests. Is it all depends on how the market is defined? Now, how you defining this market, is, for example, YouTube the competitor or is it just other streaming companies?
You know?
I tend to come at it from how do consumers actually use these technologies? What would they reach for as a substitute if Netflix weren't there or Warner Brothers streaming services weren't there. How do we actually behave I think is the most common sense way to approach that. And I think in the way that we see users sit down and they PLoP on their couch, they go to the TV. What do they put on? It's very difficult to exclude YouTube, YouTube TV from that because that's how
people use it, that's the way we're living now. So I think if regulators are trying to get to the closest version of what the real competitive effects are, you have to say, how is this actually happening with consumers and in the market. And I think it means a more broader definition of the market, which puts them more at a ten percent market share roughly. If these companies combined, if you exclude things like YouTube, and you want things
that look just exactly like Netflix. That's how you get to your thirty percent market share mark with triggers a bunch of stuff in anti trust litigation that would make it more difficult for this merger Jessica.
In antitrust, there is political theater and then there is actual process. What we showed on the screen a few moments ago and are showing now, that's political theater. Doesn't it matter really whether the DOJ takes an action here and that actually regulators are paying attention beyond the senators that post questions in that room. No, that's exactly right.
There was a lot of political haymade, and I will point out that the Senate doesn't actually have any direct jurisdiction over this merger. That's all happening at the DOJ. And if they follow traditional US anti trust laws, they're going to look at how do we define what the market is so we can think about what the market
shares involved are here. And then once you get past all of that, the question becomes, you know, on net generally, on what we can predict into the future, are there potential consumer benefits here to these two companies merging that might outweigh some of the competitive concerns that we heard mentioned yesterday in the hearing.
We've been asking this question consistently and in a process, it's probably right to ask the same question consistently, which is, is there any past precedent that we can look to that would suggest the outcome of this proposed.
Merger In the regulatory sense, I don't know it's you know,
Washington these days is setting precedent every day. But I think in the market you could look at something like the T Mobile Sprint merger and say, you know, that took two competitors into being one, but that one bigger, better, stronger competitor was able to make it three big guys on the scene alongside Verizon in AT and T instead of two, which those two individual companies at three and four wouldn't have been able to challenge to the big two.
So I think there's precedent in how things actually work in the real world, and say a lot of times mergers can be very productive, very good for the economy, for stockholders and for consumers. I think that's probably one of the more recent examples.
Consumers are the focus, though right in America at least, how are you looking at to how.
Other jurisdictions are going to analyze this deal? Because this are global companies.
Yeah, they're going to have to go within the US, They're going to have to talk to some state attorney generals and they'll have to deal with the DOJ's assessment. But you're absolutely right internationally, the standards are different here in US antitrust law. You know, it's all still the consumer harm standards. So even if there might be some concerns with labor, that's part of the discussion. But really the load star is do consumers benefit from this in
terms of maybe lower prices, more innovation, more output. There's lots of things to look at there, but that's not the case, for example, in the EU, where the focus for regulators is much more concerned with the effect on competitors to these companies and not so much just consumers.
Now, someone turned down going for the political theater, and it was David Ellison over at Paramount Skuidance.
He wasn't there. I'm sure he was tuned in.
What do you think it states in terms of him not wanting to go and address his own concerns or indeed whether he'll have his time to be able to discuss what a potential merger looks like with his business rather than that of Netflix.
I don't know.
He's probably smart to take the opportunity to not have to testify in front of Congress whenever he can. We watched those two executives yesterday or in their paychecks. But you know, the paramount stuff lingering on the sideline is sort of like a whole different bag of raccoons waiting in the wings to see how that interacts with all of this. Maybe not at all. If the board doesn't approve this, then that pretty much goes away.
But hopefully it.
Doesn't have any effect on what the board decides they want to sell, they want to do, and hopefully also doesn't have any impact on the regulatory structure. There's economics to be applied to this, there's consumers to be looked out for, and that should be a very dry, boring story that I hope we'll see sooner rather than later.
Jessica Maludim, director of the Center of Technology and Innovation at the Competitive Enterprise Institute, Thank you very much. And Video is close to investing twenty billion dollars in open Ai. That's according to sources, the deal, if completed, would be the chipmaker's single biggest investment in the chat GPT developer. Both companies declined to comment on Bloomberg's reporting. Let's get
the details with AI reporter Sharen Gafari. So, Jensen Wong has walked in front of a camera quite often recently, Sam Altman has taken to social media and also been on stage. You and I are best understanding and reporting from sources is that actually, for all the noise, this transaction's almost done. What do we need to know?
Yes, well, the key part is the twenty billion potential investment in open aiy's current funding round is almost done.
Now.
The bigger question is what about the overall up to one hundred billion commitment that in Nvidia and open Ai were supposed to agree on. They're still finalizing that agreement that is still yet to be determined.
Trian, did everything just get very confused, because at no point did it feel like in what was originally one hundred billion dollar up to agreement on paper that was going to be in exchange for Giga what's coming online was in any way sort of also bounded around the actual funding round of open Ai and can you just make a clear line of what we expected in Video and Open AI's relationship to be like from that September announcement.
Look, I think what we're seeing in these larger and larger mega funding rounds that both open Aiyanthropic and others are doing is that strategic investors are increasingly getting involved. Right, so we know as expected that it's not just going to bebc's, it's also going to be these large tech companies like in Vidia who are investing in rounds. Twenty billion is a lot for Invidia, but obviously up to
one hundred billion is a lot more. When they made that announcement of up to one hundred billion, they didn't say over what time period. They said it was a tentative agreement that has yet to be finalized, and they also said that it would be contingent upon certain milestones around how Opening Eye deploys it's chips, it's compute power,
so there's still details to be sorted out. I think what's sort of confusing is when you know there's reporting out that the deal has yet to be finalized, that there may be tension, and then at the same time we see in Nvidia in talks and as we're reported,
finalizing this very large sum in the current round. And you know, it can be true that two things are happening at once, right, so there can be a big investment in this current round, and there can be uncertainty about how much more in Vidia might give in the future. And you know what milestones opening I may or may not hit tied to that larger investment.
While excited for the rest of your reporting on the current round and who else is involved uring Gafari, we thank you very much.
Indeed, welcome back to Bloomberg Tech. If you're just joining us. The technology picture in equities public markets, it's pretty ugly. The earning story is a big part of that. AMD is now on track for its biggest drop since twenty seventeen. But there are other stories out there, right Kroen.
There are. The software sell off is a key one. It's continuing with the sector ready trailing the S and P five hundred as as investors reassess valuations and business models in the face of this rapid advance in AI. That the move has been accelerated course by Anthropic introducing a new AI automation tool fueling disruption fears across industries now here to discuss the tech market impact in both
software but also in hardware. It's handling load and portfolio manager Ude Taruvu And in fact, a lot of this software names that you analyze are balancing today. I'm looking at Adobe for example, Microsoft which is all things to all men. We've also got the likes of SAP that just managed to push up a little bit today for.
This software pain. Yeah, what do you make of it?
Look, I think there's a narrative out there that AI is going to eat up software, that AI is going to replace software, and that's dominating right now, and in a sense that investors are putting high probability to that. What we look at as long term investors is say that there's a range of outcomes that are possible. AI could be complementary and supplementary to software, not just replacing software.
And that's how you have to think about it. Right now, in a market where everything's blowing up, it looks like, hey, there's pain coming left, right and center. You step back and think about what are the sort of companies that can go through this sort of environment, go through this transformation and technology and still come out to the other side holding their own And in that sense, Microsoft is one of those companies that we think can do that.
What have you added to get that comfort, because there's much to say that not many of the seats that are actually being paid for for Copilot are being used. Yeah, I think about Adobe has been on the offensive. There's a story out today showing how they've ramped up advertising by thirty percent. Try and highlight their own AI tools. What do you want to hear and see in the data to show that these companies can weather it?
So I think before we get to the data, what we need to understand is what is the type of software that these companies are selling. Where they're points solutions, say, for example, on the legal side, you know Lexus Nexus. Maybe there's a reason why that can be displaced more by AI, which does point solutions. But Microsoft and Adobe, they do their platform solutions. What they do is not
just a function for the user. They do manage the whole workflow, the whole process of a company set up in these in companies like Microsoft or SAP, those companies from a product perspective are much more entrenched. So that's the starting point. The second point we do look at is from a data per seed, is the revenue per seed going up? And why is that going up? Is that just from pricing or increasing the number of tools that are being used? And if you look at the evidence,
there seems to be an improvement for Microsoft. The copilot is going up. Now. Is the pace of that going up differently to cloud Cowork and what they're announcing It would be because one is an enterprise product and one is much more an experimental product right now, so the growth rates would be quite different. So that's what we're looking at the data points and the number of users, percentage of take up in terms of new products.
And tools oday.
It's great to have you on the program at this moment in time, in particular, Microsoft, you know is a name that you hold in the portfolio. It's probably worth just looking back a little bit and asking what you learned from that earning sprint, because of course the market reaction in the moment was pretty severe. But what is the bigger picture for you?
The bigger picture for us is the fact that Azure is continuing to grow. Yes, it missed the bogie of you know forty percent growth, it came at thirty eight percent. To us, that matters much less than the fact that the reason that they missed is because their capacity constraint, and the reason why their capacity constraint is two years ago they made the decision to wait and see instead of blowing into growing their data centers just in anticipation of demand.
So that shows us.
The management mindset is that they will build for as they see demand, not just anticipating demand, and that's why the growth is constrained, and that's why their growth for the next quarter is constrained. So to us, that's not the end all and be ald, So therefore, missing that small number is not a bad thing. But what we do like is the fact that on the office side, on the commercial side, growth remains very strong. The part that missed was their personal computing and that's got to
do with memory pricing. Now, when you come back to hardware, huge memory price increases, there's going to be a demand destruction and that's what Microsoft are seeing. So to us, put all that together, this doesn't change the long term valuation picture. The fact that they continue to allow other llms and other models into there as you're into their asial infrastructure, is that they benefit from the overall growth in the AI ecosystem.
Bloomberg's reporting that Innvidia is close to confirming its twenty billion dollar contribution to the open Ai round that's ongoing. That of course comes from sources, but that's my best understanding. Is a portfolio manager that holds in VideA and looks in VideA, how much do you care about that story in Vidia's financial investment in open ai.
The financial investment is this quantum of that financial investment is not that important. It's what's the relationship that they're building up with open ai and how they're driving open ai in terms of their usage of Nvidia, Because the long term story for Nvidia is how much of your chips are going to be used in the next generation of training models. How much is going to go to the assets that the companies like Google and all are
coming up with. The more the deeper their relationship with these AI front team model companies is, the more likely they can they can you know, they design roadmap will tailor to these companies, and therefore they become more entrenched. What you want to see is companies getting more entrenched and invest in those companies.
Through data review of herding Loan. It's great to have you here on Blomberg Tech. Thank you very much. So coming up, Amazon's former worldwide consumer chief Dave Clark joins us. Talk about is new supply chain software startup AUGA that's next. This is Blomberg Tech.
Meta Reality Labs.
Well, it's just selected to startup Auger to provide the operating system for the company supply chain. Now, the startup was founded by the former head of Amazon supply chain itself.
Dave Clark, O a CEO. Founder joins us.
Now, you're co founder along with your partner, and I'm interested, Dave, is to what this is going to mean for all those that want the latest, greatest hardware from Meta.
How much quickly will it be in their hands?
Well, good morning, thanks for having me on. We're excited to be working with Meta in this way. We're in the midst of deploying software. Now we'll deploy with them over Q one, Q two of this year and finalizing Q three.
Interestingly, how a little spa on your website? Mess is not the only brand that I saw. Fanatics is another one. How are you scaling new companies to come on and basically be the first customers that you have for this.
Startup, Well, we started building out that. We've been building out the products for the last year. We deployed in proof of concepts with customers in the middle of last year, including Meta and Fanatics, and so what we've been doing is building the platform for how you're going to build an autonomous operating system for supply chain. At its core is really the contextual infrastructure for data across the supply
chain world that makes it happen. And because we build it as this platform of infrastructure, it's very scalable from customer to customer.
Dave, without trivializing it. Why the name auga what does it represent? I think that might give us an insight into in what the company is and what it does.
Well, if you think about the big challenge in supply chain is really trying to connect all of this disconnected data together to make it reason together, to create a layer of contextualization and understanding. And to do that you really have to have both the technology and the knowledge to drill down into an organization, to drill into the process of how they were to enable that. And so Augier fit a lot of, you know, things that we really liked and had some meaning behind it as well,
for you know, how to think about it. Also, supply chains of contacts fort like this is not you know, the world wakes up every day trying to recavoc on supply chain planners around the world, and so we thought it was it was a framing that sort of represented the grit and determination and sort of just grind through it that represents operators around the world.
We're showing video of how the software the platform works. We know you from from your time at Amazon, right and your your experience of global supply chains. You had a brief ten year at Flexport, and maybe we'll get into that another time, But I don't know you as a software guy. How did you build this platform? What are the people behind the core competency here? Please?
Yeah, well, I mean if you think about I mean, Amazon is in many ways the supply chain technology company at its core. You know, we had my team had thousands of software developers, you know, the last mile technology teams, the robotics tms, the inventory forecasting, email and neural network teams.
You know.
So well, I'm not I don't write code for a living. I've been leading and working with software teams for two decades now, and we have a phenomenal team of people, you know, many of whom were from folks I've worked with that in my past life, and some new folks
that have joined us from other places. But the unifying bit of the whole group is people incredibly passionate about customers, incredibly curious about the technology that's coming out in the world, and incredibly fast at delivering output for the people we're working for.
It's an agentic technology that's going to drive us forward in many ways, Dave. But what's so interesting is I see the retail benefits, I see the e commerce lived experience, and I can see what's happening with meta fanatics. But how does this expand supply chains is something we talk about in every type of industry right now, and particularly in the national security context.
That's right, I mean, at the core of this thing is really while agentics is the subject of so much of the conversation, the way you make agentics work is through context. Context is king, and our core platform is really about context enablement across large global enterprise data sets, unifying thousands of tables into one operable platform of data that agents can actually then work with for autonomy and
so that's extensible in all aspects of the world. Our early customers today are focused more in the consumer space, but we're also partnering looking at working with governments to enable processful You can imagine the kind of work that needs to be done to connect sovereigns, not just companies, and we think that's a big expansion area for us as we move forward in the next month months.
Taking a step back, we are currently in a day of trade where anxiety runs high on AI disruption of software companies, and I'm interested as to how you felt that ALGA is something you needed to build for yourself as an operating system rather than companies plugging in what is currently being developed by large language models.
Well, you know, as I said, I think the large language models can only do so much. You know, they're large language models are fantastic, but they're limited by the context you provide them. And what was missing, what I saw nobody building was the contextual infrastructure to make supply chains work as a unit. And so we felt, you know, that we had the right technology background, We had the right operational background, sort of right team, right technology, right
time to go out and do this. We think it's one of the you know, sort of few big enablement plays at a global scale to really you can really go out in software and change the way the physical world interacts.
Dave Clark, CEO of Auga, Great to have you here on Bloomberg Tech. Thank you very much. Let's get to the world of inference. Positron ai has raised two hundred and thirty million dollars in a Series B funding round, valuing the company at over a billion dollars. The AI chip startup is working to compete with Nvidia by developing energy efficient chips for AI inference. Joining us to discuss mtsha Agrou Positron CEO, it is good to see you.
Thank you for coming on the show. We have to say we have a new inference chip name on the show on a weekly basis at the moment, Opus right through to the next generation of accelerator for inference. We'll get to the money. But what is the Positron elevator page. What is it you're doing differently that really would allow you to take on nvideor in the market for inference.
Yeah, it's as you can, focused on memory architecture right in for inference, it's barely memory bottleneck both on memory speeds and memory capacity. Positron has deployed and you know, our first generation product already on our building our second generation, which is very much focused on building the memory closer to the systolic array, so very very fast on decode speeds, and then second is having massive amounts of memory capacity
directly attached to the chip. Our second generation will be the world's first terrabyte memory chip, going up to two point three terabytes of attached memory.
You know.
For a comparison and video Rubin will launch with three eighty four gigabytes end of this year.
So that's that's where we are really focused on.
If you think about decode application you know, video generation, code generation and reasoning models, that's the current bottleneck, very large models as well, you.
Know miteesh this issue like latency for one of a better expression, you know in video is looking the giant that is in video is looking at optics as well within the server design to overcome that. They've got a lot of cash to do it. You now have two hundred and thirty million dollars. You know, how is that going to help you progress it?
Yeah? I think Look, obviously you can't do cash tra cash comparison. I think the research to directions are different. I think that's where our technology is completely different. Well and video is looking at optics to scale out or a rack scale data center scale because they have to connect multiple chips because they're limited on memory. We are kind of bringing the scale out a little bit invert We are saying that by having that much memory, you still have to do scale out when the models are
going to be tentral in parameters long or large. But the point is you have to do it less so you can consume less chips, you consume less power and so on. Right, So it's a different type of architecture than what NVIDA is focused on. I really do believe that, you know, in the world of inference, you're going to see you know, the world you've heard the word disaggregation.
You know obviously that's going to come through.
But you know, each silicon will have Our silicon architecture will have it's kind of niche in terms of what applications it really drives through fundamentally more efficiently.
It's like the Perido curve.
You know, you are trying to minimize power consumption, You're trying to minimize energy, you're trying to minimize the cost, you know, with a per token, per video generation, and you're trying to increase the.
Speed of output, right, and each.
Of the silicon architectures will find its own niche there and we are right in that in that space.
It's only each finance.
I mean, I'm looking at Jump Trading being one of your investors and is that sort of a strategic investor. Where do you think that you will be first shipping too?
Yeah, I mean actually Jump is one of our customers for the first gen product, and it's that's the validation that you know, really kind of encouraged us to go out and raise this round.
You know, Jump became a.
Customer first, they looked at our version, they used it, they looked at our second gen roadmap, and they said, look, we got to invest in this company. And that's kind of how they came as co lead on the round. So you Trading is obviously using a lot of language models,
transformer models, or aggresive models for their applications. They're exploring that it's R and D and expanding, but they're exploring that and that's kind of where having driving lower latency or a lotting like very large context length can really make an impact for them. So that's how they got interested. From a strategic point of Caroline, you mentioned we have ARM and Qia, So it's not just customers in finance
or certain types of customers. It's the ecosystem players and you know, different types of sovereign and data center players that are getting interested in this space as they look at heterogeneous deployments.
Offsolican for inference for clothes, So.
We saw Grock.
It's focus on inference and it's interesting purchase by Nvidia. How do you see your growth now you've got money, it's about hiring talent and getting and shipping the systems. But with ARM on the cap table, do you see some sort of purchase in your future?
What do you want for the business?
I think our main focus is how can we deliver our maximum of chifts in the world, right I think that's like really what drives us fundamentally is how can we deploy millions and millions of chips. I mean we are starting at you know, started obviously with just the first prototype. We are in Balabans right now, so it's still a long way to go and you know, and videos the company and the millions, and that's kind of where we have to reach.
Everything that goes.
By the side, you know, acquisitions, you know, going public, all those things are kind of by products off the actual technical product and the consumer and the customer. From a customer point of view, I'll answer you, Carolyn, I think from from from our perspective. You know, over eighty percent of walls inference gets driven through hyperscillus sort associated customers, and that's kind of where we are really aiming to go after from specific workloads.
To to them Mintosha grow Well a Positron. Great to have you on today. Alphabet Journeys, they come out after the bell. We want to hear about what to expect, the bloombo's comment of Royo, and it's all going to be about Cloud, about capital expenditure again.
Yes exactly. I think everyone's expecting a great quarter for Alphabet. Again, it's going to be a lot of attention on Google Cloud given that Alphabet basically released Jimini three during the last quarter. So we're going to talk about how it's been rolling it out in all the products, how the distribution is going, and basically they're still dominating search, which is expected that they are and how they're depending that position against other chatbots.
Bloombo's Common Royal will be at her desk prime for alphabet post market. There are many others reporting. Qualcom is one carro Frank reality of it biggest maker smartphone processes. We'll learn about the health of the smartphone market, what's going on with memory supply chain and general demand. Of course, you know you spoke to Christiano at CES right. They want to diversify into data center, PC, auto's networking all that, but that's their bread.
And butter and robot as well.
Really, the future drivers for this business is something I'm sure they'll be trying to lean upon. But in the here and now, as you say, what is the ramifications of memory prices on some of the other clients that they serve, how are we seeing Qualcom brace for what has been just an extraordinary sell off that we're seeing into the hardware ecosystem as well. But this is as their fiscal first quarter as the one to look at. Look,
we've also got chip design in the spotlight, right. Maybe one day it wants to be a chip maker too, but right now it's a designer and ARM just up nine tens percent ahead.
Of its earnings.
How are you thinking about the strength of its own design Visa Vida competition right now?
Yeah, Jensen Wong and Video CEO changed the game for ARM based CPU. He said it just a couple of weeks ago. They're coming to market with their CPU as a standalone product. It's based on ARM architecture. AR might want to talk that up because that really did send some ripples in the market. CPU right now in demand.
Yeah, we're thinking of total revenue about one and quarter billion dollars EPs estemate nineteen cents. But let's just focus on these markets today because that does it. With this edition of Bloomberg Tech Head, we are seeing such a significant sell off across the software names, but in risk assets more broadly. Look Bitcoin seventy three thousand, where.
We stand YEP at his lowest level since the November twenty four presidential election. We have a podcast you should listen to it. This is Bloomberg Tech