CoreWeave, Supermicro Lead Tech Stocks Higher on Results

19:48Ed Ludlow:Michelle Weaver of Morgan Stanley, U.S. Thematic Research. It's great to have you on the show. Thank you very much indeed. Some news. AI startup Cognition is in early talks with investors for a new funding round that could boost the firm's valuation to$40 billion. Sources share that investors are eyeing Cognition less than three months after the firm raised a billion dollars at a$26 billion valuation. Cognition is expected to raise over a billion in this next round. Coming up, Google unveils its new Pixel 11 lineup. We'll break down the biggest announcements from its Made by Google event. This is Bloomberg Tech.

20:27The people who seem to get more done than everyone else. They're not working longer hours or running on more caffeine. They've just stopped wasting time on the stuff that doesn't move work forward. Switching apps, re-explaining context, hunting for files. Those aren't small inefficiencies. They're hours wasted every week. Superhuman Go gives you those hours back. From the makers of Grammarly, Go is an AI assistant that sits inside every tab and tool you already use. Always available and already aware of what you're working on. Ask it to draft something, summarize a long thread, pull up a file, or prep you for a meeting.

21:08Go handles it without you ever leaving the page you're on. This is what it looks like when AI actually fits into your work, instead of adding to it. It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com.

21:35Ed Ludlow:It's time for Talking Tech, and it's all about breakups. First up, Chinese AI startup Manus will resume independent operations as it part ways with Meta. The Facebook parent had planned a landmark$2 billion takeover of the AI agent developer, but Beijing forced the pair to unwind the deal over concerns it would lose control of valuable tech to a geopolitical rival. Plus, OpenAI executive Brad Lightcap is leaving the company to start something new. Lightcap joined OpenAI in 2018 and helped build its enterprise business. A few months ago, he shifted out of his role as chief operating officer and began leading special projects.

22:12Ed Ludlow:And Uber has divested from serve robotics after clashing over the deployment of delivery robots. The move underscores growing challenges in Uber's bid to become a platform for autonomous vehicles. Last month, Uber said its exclusive arrangement with robotaxi firm Waymo would end in early 2028. OK, Google is holding its Made by Google event today, unveiling a slew of new hardware, including its latest generation of Pixel smartphones. Bloomberg Senior Tech Editor, Dani Warman, is with us. What I see is prices going up and memory allocation coming down. Is that fair? That's true. And at least Google was honest about it, really cited in its statement a, quote, severe memory shortage, which, of course, has been afflicting so many consumer tech companies.

22:54And the company pledged to out-engineer, essentially, the memory shortage. But that's where we are right now, higher prices and less memory with the starting models.

23:04Ed Ludlow:OK, so we're showing the 11 lineup here. What is different from prior generation? What has incrementally changed? So things like performance, the durability. I think the most standout feature is something called highlight on one of the higher end phones, which is essentially repurposing the camera flash as a notification indicator. So that could potentially be neat and useful for users. But a lot of the improvements are otherwise sort of quality of life things. This is not a flashy new design. The company is now in its third year of what is essentially the same design. And you might compare that with other companies.

23:39Apple, for instance, the new iPhone 18 series would be in its second year, according to our reporting. And other companies like Samsung have tried bolder designs with their foldables and other flagship phones this year.

23:50Ed Ludlow:Just real quick, there's some news in wearables where we get insulin resistance features. But let's be clear, this is not continuous glucose monitoring. What is it? So this is insulin resistance is a precursor to things like prediabetes and type 2 diabetes. And what Google is doing is using a combination of AI and other kinds of biometric sensing. And it's basically tracking physiological markers that it believes are indicators of insulin resistance. So it's using AI and these other sensors like skin temperature, heart-related sensors, to paint a picture, an estimate of where your insulin resistance is.

24:32And if it senses shifts in that pattern, it sends the user a notification, at which point they can follow up with a doctor. But the company is really clear that it's not trying to offer medical diagnoses or advice.

24:43Ed Ludlow:The inverse consumer tech editor, Dana Warman. That was awesome. Thanks. Now, coming up, the CEO of Silicon Data, Carmen Lee, joins us to talk about the firm's latest funding round. It's working to build independent benchmarks for AI compute. It's halftime. And from San Francisco, this is Bloomberg Tech.

25:08Ed Ludlow:Welcome back to Bloomberg Tech. Neo clouds and AI infrastructure is really what's driving the market higher. Look at the NASDAQ 100. And the biggest percentage gain is our core weave and Nebius or elsewhere, Supermicro, up 12%, 13%. All about strong earnings where it was basically the guy for the current period that came in so significantly in both cases above the street's expectations. Let's get a digest with Bluebeck Equities reporter Ryan Vlaselica. I also point out, and I did earlier in the show, right, you know, NVIDIA is pushing higher. NVIDIA highly analogous to those two names at the neocloud level and the server level.

25:41Ed Ludlow:What else are you seeing in the markets and how big a driver are these two earnings prints today? Well, I think these two earnings reports are really reinforcing how the AI trade, especially the AI infrastructure trade, remains very strong. And that is helping to kind of improve sentiment across a lot of the tech space right now. Core, we've given a very strong forecast, as did Nebius earlier this morning, really highlighting how much demand there is for AI-related cloud computing services. And then on the hardware side, you had an extremely robust forecast from Supermicro, really underlining in a different way how much demand there is for servers, for chips, all the different kind of physical infrastructure.

26:22And I'd also flag Lumentum, the optical company. That's been a real area of focus for us lately. Really, we are seeing strength across the board for AI infrastructure.

26:32Ed Ludlow:Famous Ryan Vlaselica with The Digest. Thank you very much indeed. I want to go to private markets. Silicon Data, which provides pricing and performance benchmarks for AI compute, has raised a$30.5 million Series A to help grow its tools and capabilities. Joining us is Silicon Data CEO Carmen Lee. She's also the former head of Strategic Alliances for Enterprise Data at Bloomberg. Congrats on the round. Really interesting group of investors that have backed you. You've grown very fast. And I think that the best place to start is how are you going to use the funds? What is it you're trying to get going on?

27:07Thank you. Thank you, Ed. Thanks for having me on the show. So it's a quite exciting week for Silicon Data. As the video story rolled out, as CME announcement came out yesterday, our goal really become the independent referee for the whole compute stack. So as independent referee, we need to keep developing indices as there are more chips, there's more tokens, there's more LIM models coming online. We need to design next generation of benchmarking services. This is all going to be used to design products and software applications.

27:45Ed Ludlow:Carmen, very quickly, the valuation wasn't disclosed. Can I just ask you what the approximate valuation of silicon data was in this round? It is a very healthy range. Are we at unicorn status yet? Not quite yet, but I will definitely ping you. We'll get that out of the way. Really interesting. Gavin Baker with Valor through the Valor Atreides AI Fund led this round. What should we take from that? Beyond just the check and the capital, what are you able to get from their expertise and background here? so Gavin and the team they are the pioneer in the whole compute industry they're the one backed a lot of new clouds from very early stage and we're also backed by CME F-prime which is part of the Fidelity ecosystem Vanak obviously DRW geometry trading so we as independent referee We want to make sure our product is very well aligned and is useful for the market participants who are the natural long and natural shorts.

28:58So that's why we're excited about the whole suite of investors that's going to be heavily our clients at the same time.

29:06Ed Ludlow:Right. Right. The big thing that's to come is CME planning GPU futures that are settled against your benchmark. What's the progress there? And what is it that actually will be traded? You know, just explain to the Bloomberg Tech audience that might not be familiar with this story the basics. So it's quite exciting. So the news came out, CME, the plan date is October 5th. The first suite of indices is going to be cash settled against our H100 and H200 new cloud on-demand indices. So what's interesting, there's two actually volatility the market want to hedge. Number one is if you are natural loans of GQs, meaning you have your own tons of servers, obviously your revenue is tied with the rental rates.

29:58And for you to hedge out future volatility, you really want to show your futures. You do not care about physical delivery because you already have your servers. all you're looking to do is hedge out the price volatility in the future years. If you are a consumer of compute, consumer of GPU, obviously you're paying the rental prices. Ideally, you will lock prices in for a duration of time so you can loan the futures and lock in certain prices. So that's really the use cases for natural markets.

30:31Ed Ludlow:That takes us to the bigger picture. you referenced it, the NVIDIA$500 billion with all of those investment firms raised a lot of questions. For example, well, are these not assets that depreciate? How do you, in your benchmark, reflect computers collateral? Just go through some of those ideas. Yeah, so it's actually the same story. The NVIDIA$500 billion story is about a financing layer. CME and silicon data story is about risk management layer, right? So you need both. You can't have a market about half a trillion size without a way for the investor, the people with exposures, having a place for them to hedge.

31:18Right. And then have prices covered.

31:20Ed Ludlow:Let me ask you this, though, Carmen. Let me ask you this as a quick follow-up, and I'm sorry to interrupt. What does your data tell you about how quickly a GPU generation can depreciate? Like what I see in the news right now is A100's very high utilization, very high pricing, but they're generations old. So great, great observation. So what's interesting is A100 and H100 prices have been pretty stable for the past 20 days. The price went up about 20 % for both chips since January this year. The last 20 days has been pretty stable. The way I will read is, so there's a few points you raised before.

32:04Just because something depreciates, server is a machine, right? The machine depreciates just like aircraft carriers, just like ships, right? Everything depreciates in terms of its lifespan. Doesn't translate to just because something depreciates, meaning they have no economic value, right? So you look at residual value population, for any asset class, it's looking at future cash flow of that particular asset. And for the A100, to your point, it's an older chip. But people can still use that for different workflows and for smaller model use cases. Then they can keep charging at whatever price they supply the man-car fee.

32:42Exactly. And even L40s, right? Forget about A100. And L40s has machine learning use cases, which is a completely old chip.

32:50Ed Ludlow:That takes us back many generations. To end, I've got a question from the audience quickly, very interesting, Ducan on X. What are three signals in silicon data's data over the next two to three years that would tell you AI infrastructure oversupply has begun? So, great question. So there are three data points I can actually, which is really good to take a look. Number one is spot prices, right? Spot prices is where supply demand means. If spot prices keeps coming down for a period of time, you can argue either there's less demand or there's just outpacing supply, which is not really the case right now.

33:27Number two is forward curve, right? If the forward curve is contangled, which is kind of what we have now, that translates to people are paying even higher price to lock in a longer term contracts versus short term rental. So that doesn't say it's oversupply, right? However, if the forward curve became a very steep downward sloping curve, then the implied rate for the future, meaning it will calms down quite a bit. The third indicator is residual value. If the secondary market transactions of any servers, those prices came down, that represents people's expectation of those future revenue of those servers will be coming down.

34:08So residual value from the secondary transactions is also a good point.

34:13Ed Ludlow:Really quick, contango where future prices of an asset currently higher than their spot price. Silicon Data CEO, Carmen Lee, thank you so much. Just a note before we end, Silicon Data's index is based on historical data. It recalibrates based on the importance of each factor at the time. Bloomberg LP, which publishes the index, is the parent of Bloomberg News. Now coming up, we'll take a look at how private markets are shifting as AI gets more integrated into the workforce. Foundation Capital general partner, Joanne Chen, joins us next. This is Bloomberg Tech.

34:50Ed Ludlow:Global venture capital firm, Excel, just raised$3.5 billion to invest in early stage companies globally, raising four separate funds dedicated to backing younger startups. The firm behind AI giants like Anthropic, Cursor, Perplexity said one of the funds includes a$1.35 billion expansion and fund targeting firms specifically in the US, Europe, Israel, and India. There's an abundance of capital flowing into the latest AI startups. Our next guest comes from a firm that has a 15 year history of investing in AI. And she says the opportunity has evolved to become more exciting and important than ever before.

35:27Ed Ludlow:Foundation Capital General Partner Joanne Chen joins us now. You know, venture capital supply is there for these founders and for these startups. I guess what was more interesting at this juncture is like the decision to not invest in something, to be disciplined. How real is that struggle? Very, very real. Thanks for having me, by the way. It's great. We are a very early stage venture capital firm. And early stage means different things for different people. It does. For us, it means investing in founders that have an idea, that have a lot of ambition. They don't necessarily have customers or a product.

36:03Ed Ludlow:You don't want to see the business plan necessarily. We're investing in people at the very end of the day. And we have done this over and over again over the last 31 years of the firm's history. One of the examples is Cerebras, which went public about three months ago, which we actually incubated in our offices back in Mellon Park. And at the time, Cerebras had just a few co-founders, an idea and a ambition and a prediction that the AI market was going to explode. that maybe it's not the same field of technology, but like an interesting example that I find is humans. And so let's call it a Neo lab or something like that.

36:42Ed Ludlow:It was just four founders able to raise hundreds of millions of dollars at multi-billion valuation out of the gate. Right. And that's an interesting phenomenon because now we have examples of very large outcomes, right? As AI has changed and become much more powerful as a technology, there is an impact to organizations that could create a lot more revenue. eventually. And so investors are looking at OpenAI and Anthropic as examples of, hey, maybe we can invest at these valuations and have them become this large. And therefore, the entry valuations or the appetite for these entry valuations have gone way up for some of the types of companies.

37:18I think what's dangerous, though, is the neolabs don't have commercial traction and possibly not for a long time or for even forever. And I suspect we're going to see that play out over the years. Some are going to have very large scalability and others may end up not being able to raise capital in the future.

37:38Ed Ludlow:So how does one do due diligence on an individual or a few individuals who just have an idea? Yeah, it's not dissimilar from how do you recruit someone awesome, right? You look for exceptional people, right? If you think about the Bay Area or even San Francisco, there are fewer than a million people in San Francisco. And yes, so many companies are started here, we look for exceptional people, exceptional founders with big dreams. Oftentimes they're technical. Oftentimes they know how to sell. Oftentimes they have a very strong growth mindset because as the companies grow, the job of the CEO changes and the founder has to be able to become a CEO and also evolve as a CEO.

38:16So we look for exceptional people at the end of the day with the understanding that they will become someone different. It's hard to look for exceptionalism. There's lots of ways to do that. and for each of us, we have different ideas of what that looks like for us and I think this diversity actually gets us a really interesting portfolio of different types of backgrounds.

38:35Ed Ludlow:I'm really interested in how you're investing for the world you foresee. You published this in March. I'll get shouted out for doing that. The Great Reorg, A Human's Guide and the conclusion of the 16 pages is that AI won't simply replace people but it will reorganize companies And as part of that, you paint a picture of a world where there are smaller teams, fewer specialists, AI agents doing a lot more work. That seems a little contrarian to me because it's still net net, means less people. I would disagree with that because I think what AI is doing today is removing bottlenecks in organizations.

39:14And in order to remove bottlenecks, we do need possibly fewer teams to start with. However, if you're an ambitious entrepreneur and you're thinking, hey, I can build a billion dollar company with 100 people. Why not hire 500 people and build a much larger company? Right. Right. It doesn't stop there. I think the misconception is that AI is just about job automation or job loss. In fact, I think what's going to happen is AI is going to create a lot of use cases that we haven't yet thought about. Think about the early days of mobile. I think the first mobile app that I used was this beer drinking app where I could till my iPhone and the beer would come this way.

39:55And yet five years later, Uber came about and it really changed our words. And I suspect we're going to see the same things with AI use cases that are going to be transformative to us.

40:07Ed Ludlow:So you talk about high agency, basically being able to take an idea and actually do something with it, build, execute it, rather than sort of debate the idea. Do you have some sort of evidence of that in the portfolio since March or otherwise? One of that is that we're seeing younger and younger founders build big companies. And I think that's possible because AI can teach someone who knows nothing about a subject and get them up to speed very quickly. It can do things like guide how you're going to fundraise, come up with your marketing launch plan, help you with interview questions, help you prep for a Bloomberg TV interview.

40:46Ed Ludlow:So this is interesting because away from the investor side, the operator side of being a venture capitalist, you would pitch those services to a founder as well. We do. And we help with founders with very specific things to help them go through the zero to one journey. You can imagine having no customers getting your first is a big deal or not knowing how to fundraise and doing that for the first time. And we're also leveraging AI within our organization. Foundation Capital General Partner, Joanne Chen, it's been great to have you here with us in the studio. Thank you so much. Tech. Now, coming up, investors are betting that the AI music startup Suno will be the next Spotify, but it's got accusations of theft and AI slop to overcome.

41:24Ed Ludlow:We have that story next. This is Bloomberg Tech.

41:38Ed Ludlow:Blackmail. That's what California Attorney General Rob Bonta is calling reports. that Paramount Skydance plans to leave the state if he doesn't drop a lawsuit challenging its merger with Warner Bros. A source says Paramount's board has approved a move from California as early as October, and CEO David Ellison told executives that he's putting together a five-year plan that would shift most of the film and TV studios' jobs to the new home. AI music startup Suno's setting out to change entertainment with an app that lets anyone generate a song using simple text prompts. But the company, valued at more than$5 billion, is facing headwinds, including lawsuits from two major music labels and concerns that it could flood the Internet with AI slop.

42:22Ed Ludlow:Bloomberg Screen Time editor Lucas Shaw wrote about this in the latest big take in Bloomberg Business Week magazine. And I find this really interesting because the tagline is that investors are betting Suno will be the next Spotify. That's worth an explanation. What do we mean by that? Well, you could pick Spotify, you could pick Instagram, TikTok, any major sort of consumer internet app. That's what people see Suno as, right? They believe that there will be a winner in AI music creation consumption. And what Suno is pitching, at least, is it will change entertainment, much as Instagram. I guess the simplest way of putting it is not everyone is a professional photographer, but everyone uses Instagram either to post photos of their own or consume photos and increasingly videos that other people put up there, right?

43:07And in the case of Suno, not everyone is going to be a professional musician, but maybe anyone can make a little song and share it with friends.

43:15Ed Ludlow:So the audience is probably familiar with AI slop in the medium of video. You might see it on some of your social timelines. What is the concern or what is AI slop in the audio or music context? Well, the concern is fundamentally about dilution, right? So you You already have so many songs available to stream at your fingertips on Spotify or YouTube or pick your preferred outlet. What Suno and other apps like it do is make it so much easier to generate songs. And so the amount of songs being uploaded to Spotify and other service providers is soaring, right? There are thousands of songs all the time.

43:52Ed Ludlow:And that's what people are worried about. You speak to so many people for this Business Week story. Spotify weighs in. I think we should get to the legal dispute. Universal Music Group and Sony Music sued Suno for copyright infringement. What do we need to know about that? And I guess what's the response to our reporting around it? Well, there used to be three of the big record labels suing them. It included Warner, which has since settled. Just this morning, BMG, which is one of the largest independents, did a deal with Suno. And so we're seeing lines kind of divided in the music business among those who are willing to do a deal.

44:27and those like Sony Universal, who so far are not. I would view it primarily as negotiating posture. You know, they have a lot of history of dealing with new technology. And if the company is behaving in a way that they don't appreciate, that they think violates their copyrights, they take them to court to try to compel them to do a better deal. That is certainly what I think Universal is doing. You know, even people who work there would acknowledge that this is all sort of part of a dance and a negotiation. And the question is, how quickly can they do a deal? And the pressure on Suno would be, what if Spotify and YouTube and some of the established players come up with similar products sooner?

45:03Ed Ludlow:Lucas, you've published many stories in the last 24 hours. One is, Call Her Daddy host raises money at a$500 million valuation. The details very quickly. Well, Alex Hooper, host of one of the most popular podcasts in the world, has raised money from a firm led by kind of a former Hollywood super agent named Patrick Whitesell. It's a big number for a company that's mostly, you know, a podcast and some related videos. Bloomberg's Lucas Shaw with a must-read big take on Suno, but also a lot going on in the world of media and entertainment. Thank you very much. That does it for this edition of Bloomberg Tech.

45:39Ed Ludlow:It's an earnings-driven market right now. The NeoCloud's AI infrastructure, particularly Supermicro on the server side, is seeing a translation to upside at the NASDAQ 100, like the index level. Earnings kind of continues. NVIDIA is much later in the month. NVIDIA is one of the names with the feel good today on the idea that the infrastructure build out for AI is intact. And you see the SOX or Philadelphia Semiconductor Index outperform as a result. The Mag7 more broadly, those biggest tech names, they're under pressure. Don't forget to check out the pod, recap all the great conversations from the program.

46:14Ed Ludlow:You know where to find it from San Francisco. This is Bloomberg Tech.