Jittery Tech Markets; OpenAI Gets BofA Credit Line

24:52Designed to meet the needs of business owners at scale, this pay-in-full card elevates your travel experience and offers premium benefits and value toward business services that will take your business to the next level. Fuel your business and maximize rewards with 8x points on all purchases through Chase Travel, 3x points on social media and search engine advertising, annual partnership credits, and more. Make every journey more rewarding with a$300 annual travel credit and access to a network of airport lounges, whether you're looking for pre-flight productivity or time to rest and recharge.

25:26Chase Sapphire Reserve for Business. It's the card that gives back all you put in. Learn more at chase.com forward slash reserve business. Chase for Business. Make more of what's yours. Accounts subject to credit approval. Restrictions and limitations apply. Cards are issued by JPMorgan Chase Bank N.A., member FDIC. It's time now for Talking Tech. I'm Yahaira Anand. First up, Chinese AI firm Zipu is slated to sell$4 billion of shares after its stock stored nearly 1 ,500 % following a January listing in Hong Kong. The firm, which trades as Knowledge Atlas Technology, will use the money for research and development expansion and other investments.

26:07Plus, Samsung says it has started mass production of its most advanced AI data center storage drive designed for NVIDIA's upcoming Verirubin platform. The company says the new drive is more than twice as fast as its predecessor and is built to handle the massive data demands of AI workloads. And Meta debuted a new AI image generation system called Muse Image, its first major release since Chief AI Officer Alexander Wang took over the company's AI efforts. Muse will be integrated into Meta's apps, including Instagram and WhatsApp, allowing users to generate new images or edit existing photos. Ed?

26:47Okay, thank you, Yohaira. Coming up, it's day two of the annual Allen & Co Sun Valley Conference, also known as the Billionaires Summer Camp. Stay with us. We're going to go to the ground. We're going to get the latest details and some of the news. We're still awaiting a press conference from the President of the United States in Ankara, set to start any moment now. We will bring you the latest headlines with a focus on the situation with Iran. That's next. Stay with us. This is Bloomberg Tech.

27:26Welcome back to Bloomberg Tech. NVIDIA is cheap. The stock's down about 15 % from a May peak. sending the company's valuation to pre-AI boom levels as investors turn towards semiconductor manufacturers. Bloomberg's Carmen Reinecke, she's got the charts high. All right, thanks, Ed. Yeah, so let's take a look at NVIDIA's valuation here. So that 15 % decline in stock price has raised upon using AI to execute tasks and sharing use cases with each other. In terms of who we've seen so far, I mean, it's been a lot of the usual suspects, David Zaslav, of course, Bob Iger, Disney's new CEO, Josh Diomaro.

28:03We saw Sam Altman yesterday as well. This morning, Lachlan Murdoch. One thing that's been a little bit different this year, though, is that it seems like folks are a little bit more reluctant to come and talk to us. In years past, David Zaslav has always been very eager to get his time with the press. But not even he gave us time. We saw Comcast executives, Brian Roberts and Mike Cavanaugh yesterday. They, of course, smiled at the cameras but wouldn't come and engage. And one theory for why that's happening is a lot of these companies have deals in the pipeline or on their wish lists that are really going to hinge on getting regulatory approval.

28:39And it could be that in this very politicized environment, they're reluctant to speak out and say anything that's going to muck up their ability to get these deals across the finish line. And remind us, the deal is Paramount and Warner Brothers. What's the latest there? So the latest on the Warner Brothers Paramount deal is that they are still awaiting regulatory approval in the EU and the UK. The EU has set a deadline of July 22nd to decide whether or not they want to clear the deal or investigate further. We, of course, asked Warner Brothers CEO David Zaslav about their regulatory strategy and whether they plan to offer any concessions to the EU.

29:21but he declined to comment on that. And Comcast, of course, has also announced that it's planning to separate its NBC Universal media division from the cable unit. And if those do result in sales on either side, that will also require regulatory approval here and in other places. Bloomberg's Michelle Davis in Sun Valley, Idaho. Thank you very much indeed. Now, coming up, Eric Hippo from Lera Hippo discusses how his firm's choosing early stage startups back in a landscape where late stage companies are kind of all the rage and valuations right now kind of through the roof. We are still awaiting a press conference from President Trump in Ankara set to start any moment now and we will bring you that as it happens.

30:03Stay with us. This is Bloomberg Tech.

30:15Blue Origin is seizing on investor enthusiasm for space. It's raising$10 billion at a$130 billion valuation in its first external funding round, according to sources. The rocket company, which has until now been financed by its founder, Jeff Bezos, Blue stepping up efforts to take on rival SpaceX, though its new Glenn rocket, which can take payloads to space, is temporarily grounded following a launch pad explosion in May. Money's flowing into venture capital, but it's increasingly concentrated. PitchBook says that AI accounted for nearly 77 % of global VC deal value in the first half of this year, with more than 42 % going to just three companies, OpenAI, Anthropic, and XAI.

30:58But beneath those mega rounds, investors are becoming more selective, especially early on. Joining us to discuss that environment is Eric Hippo, managing partner at Lera Hippo, one of New York's most active early stage venture capital firms. I think there's a lot of consensus right now about what's happening, Eric. Seed stage is very hot. Valuations in some cases are extremely high. We can debate how high is too high. But it's the series A spot I want to start. You have this belief that it's tough. Why is it tough? Well, it's first of all, good morning, Ed. Good morning. It's kind of related to what you said, which is that the valuations are very high.

31:39So, you know, valuations are 50 percent, 100 percent more than they were in the past couple of years. So that's the starting point for us. Right. We're seed first investors. And so companies can raise these rounds at these crazy valuations. But then if they don't get into a high growth, fast go to market mode, they can't raise a Series A because the valuation that they started with was too high. So, you can't have it both ways. Either you start in a more modest valuation, you grow into it for your Series A, or if you don't, then the Series A investors will pass on Valuations are extremely high, as you wrote to me, in some cases, completely nonsensical.

32:28So where does the power lay? Like who's allowing a valuation at the seed stage to Series A to get to that level, the founders that are raising or those that drove and led the seed itself? I think it's the investors themselves. The investors are trying to find those big winners. And there's also, particularly on the West Coast, there's some sort of a consensus building. You know, let's pick company A to be the winner in category B. And so everybody piles on. And, you know, it's whether the company has the right moat or doesn't have the right moat. And I would argue, you know, if you're just a rapper around the foundational models, you don't have much of a moat.

33:13But nevertheless, you've been anointed as the winner. And so you raise more and more money at higher and higher valuations. And then everybody else who is not anointed the winner is kind of left on the wayside. Eric, you've been at this a long time with respect. I wondered if you could reflect on how the sort of price discovery and vetting of a founder present day is different to when you started out in your career or even just five years ago. I would say that there's way more importance, at least coming from us, in the quality of the founders. Particularly at the early stage, often it's an idea or there's a model or there's a prototype, but it hasn't really been battle-tested yet.

33:59And so the quality of the founder is really essential. and it's not just the quality of the person or these people in terms of thinking and their ability to execute, but it's also their personality and their resiliency. You know, building a business is really, really difficult and how, you know, we're going to be married to them for a long, long time and we want to make sure that they're the right people. So that really is our main focus. Now, obviously, we're interested in what they're doing in the domains. You know, we avoid investing in foundation models anything that has heavy computing layers, a lot of capex, that's really kind of for, you know, the big mega funds on the West Coast.

34:43We're looking for people who have domain expertise, who have a passionate idea, and most people obviously these days are going to be using AI to launch their product. And so we've done this now for 17 years. We're looking for that needle in the haystack, that exceptional founding team. It's so interesting because if you think the phrase seed round I've written this year, like what's the point? You can call it a coconut seed or a mango seed round. But the sort of interesting part is that a small group of largely researchers can raise such an incredibly large seed for compute. So if you're looking for the profile setup that doesn't need the compute, are they able to come on like that debut check size?

35:32What would they need it for? Well, no, look, everyone's going to use compute at some point or another. I'm looking for companies that don't have to be building the compute themselves. I see. But they'll be able to rent the compute. They'll be able to offer tokens, which are really a measurement of compute. But we're looking for people who have original ideas that are not just, again, as I mentioned, wrappers around the LLMs, around the foundational models. And we're getting to see that second generation of AI companies. And we saw this in mobile. We saw this in previous technological cycles. The first cycle is typically kind of the low-hanging fruit.

36:19I'm going to automate this. I'm going to go into customer service. I'm going to go into law and these kinds of things. The second generation are people who invent products that could not be made possible without the technology of today. And the technology of today, obviously, is AI. Those are the people that we're looking for. You're deploying your ninth seed fund. And what you told me was you're doing it with extreme discipline. So I wonder how often it's more about resisting the urge to participate in a round potentially rather than go actively chasing rounds. Yeah, look, that's where there's art coming into the science, right?

36:57So the science is, you know, can I live with a high valuation and what will it become of the company if it starts at such a high valuation? The art is, you know, do these people have taste? Do they have an idea that might not command a super high valuation today because it's brand new, it's novel, it hasn't been proven? And so, again, you weave and you bob and you go this way and our deal flow hasn't been at the highest level it's ever been. We literally are talking to thousands of companies, our entire team talks to thousands of companies every year. And so we've become more and more selective as a result of that.

37:43And we use our experience, we use our accumulated knowledge over the years to try to pick the best teams. Eric Hippo, managing partner at Lera Hippo. It's been great to have you back on Bloomberg Tech. And thank you for sharing that accrued knowledge with us. Coming up on the program, Samba Nova CEO Rodrigo Liang joins to discuss the company's latest funding. It's a billion dollars. What's driving the next wave of AI infrastructure investment, but also innovation. And we're still awaiting a press conference from President Trump in Ankara set to start any moment. We will bring it to you when it happens.

38:21This is Bloomberg Tech.

38:29Support for the show comes from Public.com. If you're actively involved in your portfolio, you probably catch yourself repeating the same actions. Buying the dip, manually sweeping idle cash, putting on a hedge. On public, you can now create AI agents that handle all these tasks on your behalf. Just describe what you want to do in plain English. Like, if the VIX hits 25, buy a put option on the S &P 500. Or, if my cash balance goes above$20 ,000, move the excess into my direct index. You approve the workflow and your agent handles the risk. Monitoring the market, watching for your conditions, and executing your strategies exactly as defined.

39:08An investing platform driven by your intent, not just your clicks. You can also get full read and write access to your account via the public API. Go to public.com slash market and fund your account in five minutes or less. That's public.com slash market. Paid for by Public Investing. Brokered services by Open to the Public Investing, Inc., Member FINRA, and SIPC. Advisory services by Public Advisors, LLC, SEC Registered Advisor. Complete disclosures available at public.com slash disclosures. When you own your own business, you own every decision. Now own the card that rewards you for it. The Chase Sapphire Reserve for Business card brings the best Sapphire Reserve benefits to business owners who expect hardworking rewards.

39:52Designed to meet the needs of business owners at scale, this pay-in-full card elevates your travel experience and offers premium benefits and value toward business services that will take your business to the next level. Fuel your business and maximize rewards with 8x points on all purchases through Chase Travel, 3x points on social media and search engine advertising, annual partnership credits, and more. Make every journey more rewarding with a$300 annual travel credit and access to a network of airport lounges, whether you're looking for pre-flight productivity or time to rest and recharge.

40:26Chase Sapphire Reserve for Business. It's the card that gives back all you put in. Learn more at chase.com forward slash reserve business. Chase for Business. Make more of what's yours. Accounts subject to credit approval. Restrictions and limitations apply. Cards are issued by JPMorgan Chase Bank N.A., member FDIC. Let's talk about healthcare for a second. It doesn't always work the way people expect it to. If you've ever waited on a prescription refill or had a hard time getting the care you needed, you know the feeling. The system should just work better for everyone. That's exactly what the people at Optum are trying to do every day.

41:01They're a healthcare company linking patient care and pharmacy services and using data and technology to drive the whole system so care is connected, not complicated, for patients and providers. Things like making it easier to get care that looks at the whole person, from primary care doctors to mental health support and even in-home care, and then using technology to make sure they all work together. Technology designed to help doctors spend less time on busy work and more time with their patients. And those prescriptions? Optum is working to bring costs down, save patients money, and make it easier to get refills.

41:35Little by little, Optum is helping make healthcare work as one for everyone. Head to business.optum.com to see how.

41:48Chip startup SambaNova has completed the first close of its Series F, raising a billion dollars at an$11 billion valuation. The company is among a growing group of startups aiming to challenge NVIDIA and supply AI infrastructure specifically for inference. SambaNova CEO Rodrigo Liang is with us with more. Let's start with the basics of what that capital allows you to do in scaling the platform? Well, look, you know, it's the inference market has broken everything open and that's where we focus. And so we're seeing this incredible demand in the market for inferencing and inferencing at really high token speeds.

42:26And so the capital allows us to really accelerate the supply chain, accelerate the build out of these racks and deliver the racks to a broad range of customers. Inference is split in two phases, pre-fill, decode. And so, you know, to the audience that might not be familiar with that, take the opportunity, Rodrigo, to explain why a specific platform for the decode phase is important, why it works better for inference. Yeah, I mean, the analogy I make is like, you know, if you're at the Disneyland ride and you have the rides that are kind of really busy and you want to organize the queues ahead of time.

42:58And so in pre-fill, what you're really able to do is organize this traffic to maximize what you need to do when the decode phase of inference shows up. And so what you're seeing is that if we use NVIDIA for the pre-filled portion and then replace that NVIDIA rack using some Manoba for decode, you're getting 2 to 3x throughput advantage on the same infrastructure and the same cost. So driving your output to a much higher level without increasing your investment. So this is the bit that I find fascinating. SM40, SM50, your proprietary tech, the way that you put it out there, five to 10 times faster on inference on the decode phase, when compared to NVIDIA GPU.

43:38But the reality is it is working alongside other accelerators on the platform, right? Explain that. Yeah, that's right. So if you look at a somewhat of a SM40, it was originally designed for the enterprise for on-prem use cases. So you could do the preview, you could do the decode altogether, and you can actually run these models at a really, really high performance, very low power, 10 kilowatts per rack versus 100 kilowatts on a typical GPU rack. Now, if you actually go at scale and you look at these cloud players, these very large frontier labs, what they're doing is actually running lots and lots of traffic behind it.

44:13And so now what you want to do is actually take NVIDIA chips, which are very good for certain things, marry them together with some of our racks, which are made for very fast decode of inference, and then getting the maximum throughput and maximizing the output of whatever power you're able to secure for the data center. Let's give some size and scope, please, Rodrigo. How many customers are actually running production workloads on SM40 or SM50 today? And I guess you'd measure that as tokens processed per day, or what would the unit of measurement be? Well, today, look, you know, Summon Over, we're in the business of moving racks.

44:52Our customers are service providers, cloud players, model builders. Those are our customers. We don't sell tokens. We really sell racks. And so if I look at kind of what we did this summit over, we're already in the dozens of customers and we'll touch triple digit by the end of the year. And so you were in the model of deploying racks, which then those customers of ours would turn into token services that allow them to offer these premium tokens, very fast tokens on the largest models, out to their developers, out to their users. JP Morgan's the big one, right? It's a really interesting case study.

45:26Explain what it is that J.P. Morgan is able to achieve with your technology. Yeah, really excited about this announcement. J.P. Morgan has selected Salmonovo to be the inference provider for the bank. And what you're seeing now is where the attention continues to be on the frontier labs, on the hyperscale clouds. One big segment of inference or AI is waking up, and that's enterprise. The enterprise, frankly, JP Morgan has always been the leader in actually using technology to drive the business. And so they're using Salmonova because they can bring this technology onto their premise, private data put into these racks, secure data completely within their own firewalls.

46:09And you can do all of this highly regulated business in production within the confines of a nicely protected environment. Rodrigo, I think you'd admit, right, there are many inference-specific platforms out there in the world at different stages in their life cycles. But, you know, I want to go back to this idea that the SM40 and SM50 work in conjunction with other accelerators, but in particular the GPU. The GPU architecture relies on high bandwidth memory. Cerebris, for example, has talked a lot about its use of SRAM, so it's not affected by the bottleneck that is HBM. How does that play out for SambaNova?

46:50Well, it's a data flow architecture which is very SRAM based. And so what it does, it is able to actually take all the benefits of the SRAM, the very, very fast SRAM, and generate the speeds. One of the benefits of having HPM is now you're able to run the big models. Where the SRAM, traditional SRAM based architectures that don't have HPM, they can't run the trillion parameter models. You have to quantize those models. And so we want to tackle the hard models. And so we use HBM. But here's what we did on Samanova. We used HBM that was N-1 technology, HBM that was already in mature production, which allows us to actually generate significantly more supply versus competing with NVIDIA on, say, the latest and greatest, newest HBM.

47:38And so we're delivering high performance using mature HBM technology where the supply is significantly better. Let's end on by going back to the round. Very interesting who's now on the cap table or increased position. What happens next for you guys? Like capital intensive on the R &D side, on the scaling side? How are you going to manage your finances going forward? Look, I mean, this is something that, you know, we're heading to scale and you see the investors coming in, you know, from General Atlantic to Capital Group, Tiro Price, Columbia Seligman, you know, just incredible investors that have great track record going to the public markets.

48:13And so we're actively driving the scale. We see incredible ramp up on our demand. And so we're just focused on delivery and driving the revenue to the point that allows us the option of actually going into public markets. Hey, Bloomberg Tech, technology, but also the money behind it. Rodrigo Yang, CEO of Samba Nova, thank you very much for your time. We're still awaiting a press conference from the president of the United States in Turkey. We will bring you that press conference and the president as soon as it starts in Ankara. There have been many headlines already this morning relating to Iran, and that is the expectation of where that will focus, as well as the discussion that has happened over a number of days relating to NATO.

48:56That is why they are there. That does it for this edition of Bloomberg Tech. The markets right now are pretty mixed and pretty all over the place. We had some really interesting insight from those in the markets, both private and public. Recap all of that on the podcast. You know where to find it. On all of the Bloomberg platforms, as well as online on Apple, Spotify, and iHeart.

49:23When you're running a business, the best days are the ones where priorities stay on track. For midsize and large companies, risk can affect multiple parts of the organization at once, from property and liability to cyber and regulatory challenges. At that level, managing risk becomes an ongoing discipline. At the Hartford, the focus is on helping businesses manage risk before it turns into something more disruptive. And when losses do happen, that work is paired with insurance coverage shaped by years of underwriting, risk engineering, and claims experience. Learn more at thehartford.com slash risk mitigation.

49:58Policies provided by Hartford Fire Insurance Company and its property and casualty affiliates, Hartford, Connecticut. As industries evolve faster than ever, companies need an environment that accelerates strategic growth, and Michigan delivers on that promise. From emerging startups to global enterprises, Michigan offers what executives value most, a resilient, innovative ecosystem, Diverse communities that attract top talent and a quality of life that supports work-life balance. With our unified Team Michigan approach, businesses scale faster and compete at the highest level. Michigan, pure opportunity.

50:31Seize your opportunity at michiganbusiness.org. These days, it seems like AI agents are just about everywhere you turn, every field and every function. But without identity, you can't trust they'll serve your business instead of jeopardizing it. Fortunately, Okta helps you get identity right by securing your AI agent's identities, giving you a single layer of control, a single standard of trust. So whether an AI agent supports a single user or your entire enterprise, with Okta, you'll turn risk into opportunity. Secure every agent. Secure any agent. Okta secures AI.