# OpenAI Unveils First Custom AI Chip With Broadcom

An episode of [Bloomberg Tech](https://invisiblenetwork.ai/podcast/1611767434), 24 Jun 2026, 47 min.

Audio: https://podtrac.com/pts/redirect.mp3/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/41764a4f-fc64-4e11-89ba-ae7c0030ab5e/bcfd636b-5d5a-47c9-8911-b4730123232c/audio.mp3?utm_source=Podcast&in_playlist=9caafc41-289c-4115-995d-ae7c0030ab75

Bloomberg’s Ed Ludlow takes a look at OpenAI's first custom AI chip, called Jalapeno, developed in partnership with Broadcom. Plus, SK Hynix is planning to raise $29 billion in a landmark US listing, and Cerebras CEO Andrew Feldman weighs in on the company's first quarterly earnings since going public. See omnystudio.com/listener for privacy information.

## Transcript

[0:00] Speaker 1: Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is live
[0:15] Speaker 1: from the heart of Silicon Valley with ed La though
[0:18] Speaker 1: in Van Francisco.
[0:22] Speaker 2: This is Bloomberg Tech coming up.
[0:24] Speaker 3: Open Ai unveils its first custom AI chip called Kalapino,
[0:28] Speaker 3: developed in partnership with Broadcom, plus s k Highnex's plane
[0:32] Speaker 3: to raise twenty nine billion dollars in a landmark US listing,
[0:36] Speaker 3: racing to increased capacity to meet memory chip demand, and
[0:40] Speaker 3: Cerebrus reports quarterly earnings for the first time since going public.
[0:44] Speaker 3: CEO Andrew Feldman joins us later this hour. Let's get
[0:49] Speaker 3: to our top story custom Silicon open ai is tackling
[0:52] Speaker 3: one of AI's biggest challenges, the supply and costs of computing.
[0:56] Speaker 3: The company unveiled Kalapino Intelligence process. So that's the wafer,
[1:01] Speaker 3: a custom AI chip developed with broad coomp which the
[1:03] Speaker 3: companies say carries a fifty percent lower cost versus a
[1:07] Speaker 3: typical AIGPU. The focus faster, cheaper AI inference shares a
[1:12] Speaker 3: broad Open pretty higher round two percent. We're off session highs,
[1:16] Speaker 3: but still up a percentage point or so. In a
[1:18] Speaker 3: market where tech at the index level kind of flat.
[1:21] Speaker 3: This is outperformance. Blinbo Tech editor Seth Figgerman is here
[1:25] Speaker 3: with the details. This is the big picture, right, open
[1:28] Speaker 3: ai is compute constrained, heavily reliant on Nvidia.
[1:32] Speaker 2: They want to diversify. What do we need to know?
[1:35] Speaker 4: Yeah, that's very I think, you know, Opening I wants
[1:38] Speaker 4: to own more of the infrastructure stack, have more flexibility here,
[1:42] Speaker 4: and then I think importantly, use the expertise that it's
[1:44] Speaker 4: developed into what makes air models run better and more
[1:47] Speaker 4: efficiently and apply that to the hardware and the same
[1:49] Speaker 4: way that I think we're seeing Google do with its TPUs,
[1:52] Speaker 4: and the hope would be that over time they can
[1:54] Speaker 4: build a chip that's more cost efficient and also better
[1:57] Speaker 4: performing and eventually cuts down its own costs.
[2:00] Speaker 3: Bloomberg News spoke to both Hocktan the CEO of Broadcom,
[2:05] Speaker 3: and open ai is hardware chief. I guess you know,
[2:08] Speaker 3: from raw Comm's perspective, there was quite a lot of
[2:09] Speaker 3: fighting talk. Right, they see a lot of demand for
[2:12] Speaker 3: a six or custom silicon. But what was interesting to
[2:15] Speaker 3: me is hock Tans saying to us they see a
[2:18] Speaker 3: world where every frontier lab goes to have a custom chip.
[2:21] Speaker 2: Why yeah, I mean.
[2:24] Speaker 4: It gets back to what we said about Openny a
[2:26] Speaker 4: minie ago, I think giving them the versatility to kind
[2:29] Speaker 4: of control their infrastructure stack and build customer hardware that
[2:32] Speaker 4: meets their needs. And now I think Hocktan's com and
[2:34] Speaker 4: he's saying there's not that many of these frontier model developers.
[2:37] Speaker 4: So I think agree between the lines we're kind of
[2:39] Speaker 4: talking about. Is anthropic, I'm going to move in this
[2:41] Speaker 4: direction and that's certainly possibility. And to your other point,
[2:44] Speaker 4: I think Hocktan sounds quite bullish generally on the prospects
[2:48] Speaker 4: for this chip and chips like it going forward. He
[2:50] Speaker 4: sees no limit to the demand for this kind of
[2:53] Speaker 4: infrastructure and is predicting that the demand and roll out
[2:57] Speaker 4: for this chip next year will exceed his prior estimates
[2:59] Speaker 4: for one point three gigawads to supply on.
[3:03] Speaker 3: The open AI side, it's the latest example of them
[3:06] Speaker 3: being prepared to spend. Did we get any sense from
[3:09] Speaker 3: our conversation with the company about how this custom chip
[3:13] Speaker 3: will rank in their overall plan for compute, whether it's leasing,
[3:17] Speaker 3: data center, buying and video chips, et cetera.
[3:20] Speaker 4: Yeah, you know, they remain pretty tight. Lop DOWNE what
[3:21] Speaker 4: the financing for this will look like. We have previously
[3:23] Speaker 4: reported that Opening I planned to spend tens of billions
[3:26] Speaker 4: of dollars on these chips, with Broadcom part of its
[3:28] Speaker 4: hundreds of billions of dollars in infrastructure commitments. Seeing Broadcom
[3:33] Speaker 4: has previously said that they were standing up a chip
[3:35] Speaker 4: financing vehicle that could help Opening I's efforts, but it's
[3:38] Speaker 4: unclear that's the primary financing mechanism for what Opening I
[3:41] Speaker 4: will do here.
[3:43] Speaker 3: Bloomberg's AI editor Seth Figemann, thank you very much. Indeed,
[3:46] Speaker 3: s k Heinez plans to raise twenty nine point four
[3:49] Speaker 3: billion dollars in its US market debut, a deal large
[3:53] Speaker 3: enough to rank among the top five share sales of
[3:55] Speaker 3: all time. The listing comes at a time when markets
[3:58] Speaker 3: are hungry for memory chips and semicindut to companies are
[4:01] Speaker 3: rushing for capital to expand capacity meet that demand. The
[4:05] Speaker 3: BOS Peter Elstrom, who leads our coverage of Asia tech,
[4:08] Speaker 3: joins us that's part of it, right, Expand memory fab
[4:12] Speaker 3: capacity raise a lot of money. There's also this idea
[4:15] Speaker 3: of like valuation and prestige of an ADR situation where
[4:20] Speaker 3: you get exposure to American investors.
[4:25] Speaker 5: Yeah, that's right.
[4:26] Speaker 6: Sk Heinez is a company we don't talk about that
[4:28] Speaker 6: often on the show. Obviously a leading memory chip company,
[4:32] Speaker 6: but they often play second fiddle to Samsung, even in
[4:34] Speaker 6: South Korea. But what we've seen from them over the
[4:37] Speaker 6: past three years is really they've come from quite far
[4:39] Speaker 6: behind to essentially the same market cap size as Samsung
[4:44] Speaker 6: at this point. And that's largely because they've been so
[4:46] Speaker 6: successful in these memory chips for AI applications that HBM
[4:50] Speaker 6: chips that we've talked about before. They especially HBM chips
[4:54] Speaker 6: are DRAM chips stacked on top of each other, and
[4:57] Speaker 6: their technology for doing that stacking has been very good.
[5:00] Speaker 2: They got Nvidia's blessing.
[5:02] Speaker 6: They've been able to jump ahead of Samsung in that
[5:05] Speaker 6: market quite effectively.
[5:06] Speaker 5: Samsung has struggled a little bit.
[5:08] Speaker 6: So now what we're seeing with this twenty nine billion
[5:10] Speaker 6: dollar fundraising is really them announcing to the US, hey,
[5:14] Speaker 6: that they're ready to raise this kind of capital. They
[5:16] Speaker 6: want more exposure in the US and yeah, they'd like
[5:19] Speaker 6: more of that pe multiple that you see with the
[5:21] Speaker 6: US tech companies too. They've traded at a discount to that.
[5:24] Speaker 6: If they get some of that ADR exposure, that may
[5:27] Speaker 6: give them a little more opportunity to increase that valuation, and.
[5:31] Speaker 3: Just on the mechanics, Peter, so it will be ADRs
[5:33] Speaker 3: I believe, on the Nasdaq. So we have a timeline
[5:36] Speaker 3: for this. And how much of a surprise was it
[5:38] Speaker 3: that sk plowed ahead with this.
[5:42] Speaker 6: Well, it's certainly they've talked about wanting to raise money
[5:45] Speaker 6: in the US for a while, so I think in
[5:48] Speaker 6: terms of direction, this has been expected. I think in
[5:51] Speaker 6: terms of magnitude, it's a bit of a surprise though
[5:53] Speaker 6: it's a lot of money. As you mentioned, it could
[5:55] Speaker 6: be one of the top five offerings of its kind. Ever,
[6:00] Speaker 6: they plan to start trading these ADRs on July tenth.
[6:03] Speaker 6: They've talked about that. They still have a couple of
[6:05] Speaker 6: steps to go before they formally do that, but that's
[6:07] Speaker 6: the plan. At this point, they'd be able to raise
[6:09] Speaker 6: that money and then, as you say, they really want
[6:11] Speaker 6: to spend this money and trying to solve this choke
[6:14] Speaker 6: point that they've had in the AI industry where memory
[6:16] Speaker 6: chips are in short supply. Prices have been soaring, which
[6:19] Speaker 6: is driving profits for a bunch of companies, including Micron,
[6:22] Speaker 6: as we'll see after the bell today. But they want
[6:25] Speaker 6: to be able to build additional capacity in South Korea.
[6:27] Speaker 6: They're also expanding in the United States, where they have
[6:30] Speaker 6: some facilities too.
[6:32] Speaker 3: Bloomberg Tech Executive Editor details from Thank you very Much.
[6:35] Speaker 3: The AI hardware story and memory in particular are set
[6:38] Speaker 3: to dominate market conversation today. We've got the perfect person
[6:41] Speaker 3: to speak about it. Selin Wu is a portfolio manager
[6:44] Speaker 3: on Lazard Asset Management's Global Robotics and Automation team. She
[6:47] Speaker 3: manages the firms tech in Ai Focus ETF, called TECHI
[6:51] Speaker 3: or TEKY and actively managed ETF where sk Heinex and
[6:56] Speaker 3: its career and shares are the top holding.
[6:58] Speaker 2: Welcome to the program, Thanks for having me.
[7:00] Speaker 3: The situation with the memory makers is fascinating.
[7:05] Speaker 2: Demand is exceeding supply.
[7:07] Speaker 3: Historically, that's an enviable position to be in, and the
[7:11] Speaker 3: shares of those companies probably reflect that. How do you
[7:14] Speaker 3: see investor appetite to be exposed to the memory trade?
[7:18] Speaker 7: Yeah, I mean the shares have been fantastic, But even
[7:21] Speaker 7: after that, I think memory stocks still love for a
[7:25] Speaker 7: very compelling investment case. And here's why. I don't think
[7:28] Speaker 7: it's necessarily just driven by temporary supply bottlenecks. Instead, I
[7:33] Speaker 7: think there's something structured going on in the cycle. For example,
[7:37] Speaker 7: memory's strate peagic value is rising and changing to a
[7:42] Speaker 7: primary driver for performance. Think about how cloudsher providers, one
[7:46] Speaker 7: of their largest clients. They're still adding more memory despite
[7:51] Speaker 7: the price of inflation because this is still the most
[7:54] Speaker 7: cost efficient way to maximize their system level performance. And
[7:59] Speaker 7: when of the when token demand, excuse me, the token
[8:02] Speaker 7: demand is rising on the back of the inference as
[8:05] Speaker 7: well as agent the AI. I think the bottlenecks in
[8:08] Speaker 7: terms of the capacity and bendwidth are becoming even more challenging,
[8:12] Speaker 7: and this is a place where memory can come in
[8:15] Speaker 7: and help navigate those challenges.
[8:18] Speaker 3: Does it matter that investors will have the opportunity to
[8:22] Speaker 3: have US listed shares ADRs of sk How does that
[8:25] Speaker 3: kind of change the mechanics of the market for you
[8:28] Speaker 3: as well?
[8:29] Speaker 7: Yeah, I mean a clearly the more exposure to more
[8:32] Speaker 7: diverse type of shareholder base is clearly a very positive
[8:36] Speaker 7: indication for any company in the world, especially when your fundamentals,
[8:41] Speaker 7: when your fundamentals are improving. So yeah, super exciting us
[8:44] Speaker 7: and looking forward to the development.
[8:46] Speaker 3: A couple of months ago, sitting down with Jensen Wong
[8:49] Speaker 3: and I said to him, do we even need the
[8:51] Speaker 3: textbooks anymore? That would tell us historically memory is cyclical,
[8:56] Speaker 3: it is boom and bust. All of the evidence suggests
[9:00] Speaker 3: that in the context of HBM going into data center,
[9:02] Speaker 3: it just doesn't behave the same way.
[9:04] Speaker 2: What is your thesis on that?
[9:05] Speaker 7: I mean, I agree with that. I mean, like I said,
[9:08] Speaker 7: there's something structural going on. Yes, there is certain type
[9:12] Speaker 7: of cyclicality going on indoor market. But at the end
[9:16] Speaker 7: of at the end of the day, what you need
[9:18] Speaker 7: to remembered is structure side of booth, demand and supply,
[9:22] Speaker 7: and think about the supply side.
[9:24] Speaker 2: Despite the fact that all the three.
[9:25] Speaker 7: Major companies are scrambling to add new supply, there are
[9:29] Speaker 7: three factors that you have to remember. Number one, manufacturing intensity.
[9:34] Speaker 7: Manufacturing complexity is rising. Number two, capital intensity to open
[9:39] Speaker 7: a new fab is getting more expensive, and number three
[9:43] Speaker 7: the difficulty of technology and migration all together pointing out
[9:47] Speaker 7: that overall supply demand balance will stay in favor of
[9:51] Speaker 7: the memory stocks. I mean, I think that's really great.
[9:53] Speaker 3: I'd love to talk about your actively managed ETF for
[9:56] Speaker 3: a moment. T e k y Techy Techy I launched
[9:59] Speaker 3: last April, write sixty million in assets currently. What I
[10:02] Speaker 3: find so interesting about the composition of it is something
[10:05] Speaker 3: we've talked about quite recently on the show. There are
[10:08] Speaker 3: the capital expenditure deployers and there are the capital expenditure
[10:11] Speaker 3: recipients in the same basket when you look at the
[10:14] Speaker 3: top holdings. Is that a conscious, active decision on composition.
[10:19] Speaker 7: I think it's the composition from our bottom up stock
[10:23] Speaker 7: selection and portfolio construction. But like you said, that's exactly
[10:26] Speaker 7: where majority of the capital is invested in currently. On
[10:30] Speaker 7: one hand, there are big spenders, big kip expenders that
[10:33] Speaker 7: are trying to build new competitive positioning in this AI
[10:38] Speaker 7: homage pace. But on the other hand, we find a
[10:41] Speaker 7: lot of exciting opportunities on the companies that we see
[10:44] Speaker 7: these capitals AI hardware supply chain, for example. I think
[10:47] Speaker 7: that's why there's more to come. In fact, we just
[10:51] Speaker 7: recently came from Asia where we sat down with a
[10:53] Speaker 7: number of different companies into supply chain. Everyone is telling
[10:57] Speaker 7: us how they are seeing an extended visibility from customers,
[11:02] Speaker 7: how there are more conversations about ltas. Everything collectively is
[11:07] Speaker 7: highlighting that demand continues to substantially exceed the supply and
[11:12] Speaker 7: the AI demand outlook remains pretty robust as well.
[11:15] Speaker 3: For in the ETF context, it's a very competitive market.
[11:19] Speaker 3: How closely are you thinking about flows? And again, I
[11:22] Speaker 3: think people give you a lot of credit Sealine for
[11:24] Speaker 3: identifying sk as being just critical to the broader infrastructure
[11:28] Speaker 3: build out right now, But you know what happens next
[11:31] Speaker 3: to your mind? How do you see the world changing
[11:34] Speaker 3: in the next I guess six months to twelve months.
[11:38] Speaker 7: I mean even more than that. I mean that's why
[11:40] Speaker 7: we focus on AI tech stack, and at the top
[11:43] Speaker 7: of the step we have application layer and this is,
[11:46] Speaker 7: for example, where we are going to expect really large
[11:50] Speaker 7: opportunities from physical AI. We think physical AI is going
[11:54] Speaker 7: to be multi trillion dollars long term opportunity. Essentially, AI
[11:59] Speaker 7: is in because it's a primary driver for productivity gains,
[12:03] Speaker 7: and historically productivity growths tend to translate to massive economic expansions.
[12:09] Speaker 7: I think the same thing is going to happen for AI.
[12:12] Speaker 7: All the innovation we're seeing today is eventually going to
[12:15] Speaker 7: open up very significant new end markets in physical AI,
[12:19] Speaker 7: such as fully autonomous transformation transportations as well as a
[12:24] Speaker 7: humanoid robot. That's why we're really excited about physical AI
[12:29] Speaker 7: and in the meantime, that being said, it will still
[12:32] Speaker 7: take a couple of years for this to turn into
[12:34] Speaker 7: actual corporate earnings. So here's what we focus on instead.
[12:38] Speaker 7: Companies with vertically interiorted the manufacturing excellence that can speak
[12:43] Speaker 7: to the market with scale advantage for example, or open
[12:46] Speaker 7: source platform that can expedite the excelevation adoption and everything.
[12:51] Speaker 7: And lastly, global technology companies that can massively benefit from
[12:56] Speaker 7: the mass adoption of the application in it sulf something
[12:59] Speaker 7: to watch is something we're still pretty excited about.
[13:02] Speaker 3: Seline Wu from Lazard Asseid Management, first time on Bloomberg Tech.
[13:06] Speaker 3: Really grateful to have you here, Thank you very much. Indeed,
[13:08] Speaker 3: now coming up, Meta and Microsoft are leading the spending
[13:12] Speaker 3: spree on future data center leases, adding tens of billions
[13:16] Speaker 3: of dollars just last quarter alone. We'll get into the
[13:18] Speaker 3: details next. This is Bloomberg Tech. Okay, today's big number,
[13:34] Speaker 3: eight hundred and fifty billion dollars. That's how much the
[13:37] Speaker 3: world's largest cloud computing companies have committed in future data
[13:41] Speaker 3: center leases. Leading the charge in the AI spending spree,
[13:45] Speaker 3: Meta and Microsoft both committing tens of billions more just
[13:48] Speaker 3: last quarter. Bloombo's Brodie Ford has an astonishing story and
[13:52] Speaker 3: a hell of a.
[13:53] Speaker 2: Chart to show us.
[13:54] Speaker 3: What I find interesting about this is that these are
[13:57] Speaker 3: future leases, essentially leases planned or announced separate from what
[14:02] Speaker 3: is already in operation, what they are already contractually committed to.
[14:06] Speaker 8: It's a really good point, right, So we're talking about
[14:09] Speaker 8: eight hundred and fifty billion on data centers, which in
[14:12] Speaker 8: many cases aren't even built yet. It's they are working
[14:16] Speaker 8: through their current projects they have, and these are the
[14:18] Speaker 8: ones that they've committed to leasing once they're operational. And
[14:22] Speaker 8: so a lot of people wonder about what's the long
[14:25] Speaker 8: term spending trajectory on data centers? How long does this last? Well,
[14:29] Speaker 8: there's eight hundred and fifty billion dollars of leases that
[14:32] Speaker 8: are going to be starting in the coming years.
[14:35] Speaker 3: So when I was reading the story, I was trying
[14:37] Speaker 3: to understand the future costs. It's an amazing chart, by
[14:41] Speaker 3: the way, we're just showing on the screen right now,
[14:43] Speaker 3: like the level of capital and it's the commitment grows.
[14:47] Speaker 3: But where did their current financial healths and balance sheet
[14:53] Speaker 3: stand relative to those future commitments.
[14:55] Speaker 2: Is there a concern about that?
[14:58] Speaker 8: I think there's a lot of concern that seem to
[15:01] Speaker 8: always ebb and flow around the ROI of all this spending, right.
[15:05] Speaker 8: I mean, it feels like every quarter or two the
[15:07] Speaker 8: general sentiment flips about whether the market's worried or not,
[15:11] Speaker 8: but every hyperscaler seems to agree that they must spend
[15:15] Speaker 8: an incredible amount to catch up an AI to have
[15:18] Speaker 8: enough data center capacity. I think about Microsoft, who through
[15:22] Speaker 8: a lot of twenty twenty five said we're nervous, we're
[15:25] Speaker 8: actually going to put on ice a lot of these leases,
[15:27] Speaker 8: and they've come to regret it pretty deeply. And I
[15:30] Speaker 8: think a lot of the biggest tech companies, as we
[15:32] Speaker 8: can see in this chart, are hoping to avoid that
[15:35] Speaker 8: kind of scenario.
[15:37] Speaker 3: Microsoft added forty one billion dollars in commitments to a
[15:39] Speaker 3: total of almost one hundred and ninety seven billion dollars.
[15:42] Speaker 3: Compare and can trust really quick Oracle.
[15:45] Speaker 8: Oracle's very interesting because they signed a ton of leases,
[15:49] Speaker 8: largely for open AI, and they've been flat for the
[15:52] Speaker 8: last two quarters, which says to us that they're digesting
[15:55] Speaker 8: a lot of capacity, contrasted with somebody like Meta, who
[15:58] Speaker 8: is just signing as much much as they seem to
[16:00] Speaker 8: be able.
[16:01] Speaker 3: To Bloombanks, Brady Ford, thank you very much. Just get
[16:04] Speaker 3: out to New York where bloom Bek Jihara and and
[16:07] Speaker 3: is standing by Hi.
[16:07] Speaker 9: Yahara, Hi ed It's time now for talking tech. First
[16:11] Speaker 9: up Softbanks. Masayoshi Sun says he has no plans to
[16:15] Speaker 9: retire anytime soon as he looks to capitalize on the
[16:18] Speaker 9: AI boom. Speaking at an annual shareholder meeting, the sixty
[16:22] Speaker 9: eight year old visionary fired back at skeptics, saying, quote,
[16:26] Speaker 9: the AI revolution has only just begun. Calling it a
[16:29] Speaker 9: bubble is an insult plus and the latest showdown between
[16:33] Speaker 9: Chinese big tech and national security. Ali Baba has sued
[16:37] Speaker 9: the US Defense Department demanding removal from its blacklists. The
[16:41] Speaker 9: e commerce giant called a designation arbitrary and unjustified, following
[16:46] Speaker 9: a US crackdown earlier this month that accused several of
[16:49] Speaker 9: China's top companies of aiding Beijing's military, and TikTok parent
[16:54] Speaker 9: by Dance is in early talks to secure a record
[16:58] Speaker 9: twenty billion dollars global loan. Well the exact use of
[17:01] Speaker 9: the proceeds remains unclear. The massive borrowing push comes as
[17:05] Speaker 9: the company wis boosting data center and AI capital.
[17:08] Speaker 2: Spending ed Thank you very much, you Hira.
[17:11] Speaker 3: Now coming up on the program defense startup, Hadrian discusses
[17:15] Speaker 3: new funding at a seven point five billion dollar valuation
[17:18] Speaker 3: that would be four times its past or latest value.
[17:22] Speaker 2: We're going to discuss that next. This is Bloomberg Tech.
[17:38] Speaker 3: Can AI replace human effort in the realm of scientific discovery?
[17:42] Speaker 2: In the latest episode of The Circuit.
[17:44] Speaker 3: Bloomberg's Emily Chang spoke to Jennifer Dowdner, the inventor of
[17:47] Speaker 3: the groundbreaking crispher gene editing technology, about the future of biology.
[17:52] Speaker 2: Listened to this.
[17:54] Speaker 10: Biology is complex. We're not going to be able to
[17:57] Speaker 10: simulate our way to an understanding of the human. We're
[18:00] Speaker 10: not going to be able to avoid the need for
[18:03] Speaker 10: certain types of testing. I do think there are opportunities
[18:06] Speaker 10: to increase the efficiency in which we make discoveries about
[18:11] Speaker 10: the way our bodies work and the way they interact
[18:14] Speaker 10: with drugs that will be effective, and I think AI
[18:16] Speaker 10: will be helpful there. But it's going to come down
[18:18] Speaker 10: to training models on the right kind of data, and
[18:21] Speaker 10: a big need for better and more data for training
[18:25] Speaker 10: models if we want to achieve that.
[18:27] Speaker 9: An Open AI executive recently suggested that if a discovery
[18:30] Speaker 9: happens on chat GPT, let's say a drug discovery that
[18:33] Speaker 9: open AI should get a cut of sales.
[18:36] Speaker 11: What do you think of that?
[18:37] Speaker 10: Good luck expand how are chatbots going to change drug discovery?
[18:42] Speaker 10: I'm not sure the answer to that yet.
[18:43] Speaker 5: I don't know.
[18:44] Speaker 10: Lots of people are, of course, very very hopeful, so
[18:47] Speaker 10: I'm very helpeful about it. But I think that innovation
[18:51] Speaker 10: is still really in the domain of human beings right now.
[18:54] Speaker 10: I'm not seeing chatbots in our own experience innovating. They
[18:58] Speaker 10: can be helpful with summer rising data, that can be
[19:01] Speaker 10: helpful in writing reports and things of that nature, but
[19:04] Speaker 10: I'm not seeing chatbots coming up with a brand new
[19:07] Speaker 10: idea for something that nobody else ever thought of.
[19:09] Speaker 2: So you're saying AI can't innovate.
[19:12] Speaker 10: I don't know if it can't innovate, I just don't
[19:14] Speaker 10: think it is right now.
[19:15] Speaker 2: What about after the AGI moment?
[19:17] Speaker 10: Well, I never say never, so maybe that'll happen, but
[19:20] Speaker 10: I'm not holding my breath.
[19:21] Speaker 3: That was Bloomberg's Emily Chang speaking with Jennifer Dowdner, Nobel
[19:25] Speaker 3: Laureate and founder of Innovative Genomics Institute. What's the four
[19:29] Speaker 3: episode on Bloomberg Television tonight at six pm Eastern or
[19:32] Speaker 3: on Bloomberg Originals at eight pm Eastern time. Defense manufacturing
[19:38] Speaker 3: start up Hadrian has been in talks to more than
[19:40] Speaker 3: quadruple its valuation in a new funding round, according to sources,
[19:44] Speaker 3: Bloomberg's venture reporter of BEC and Torrents and I teamed
[19:47] Speaker 3: up on this one.
[19:48] Speaker 2: So let's go through the basics.
[19:50] Speaker 3: Right, these are talks, they're at a certain stage, but
[19:53] Speaker 3: they only raise money right at the beginning of this
[19:55] Speaker 3: year at a certain valuation. What's the jump and what
[19:57] Speaker 3: do we need to know?
[19:58] Speaker 2: That's correct ed?
[19:59] Speaker 12: So Hadrian raise money at a one point six billion
[20:02] Speaker 12: dollar valuation just in January. They've now had conversations to
[20:05] Speaker 12: raise up to a billion dollars in this latest round.
[20:08] Speaker 12: That would more than quadruple their valuation, as you said,
[20:10] Speaker 12: to seven point five billion dollars. The investor demand on
[20:13] Speaker 12: this one is pretty significant. We reported that several existing
[20:16] Speaker 12: investors are in talks for this one. We should note
[20:20] Speaker 12: that when we reach out to comment, Hadrian said that
[20:23] Speaker 12: the information provider was incorrect decline to provide further context.
[20:26] Speaker 2: That said, we sitten by our reporting.
[20:28] Speaker 12: This is an area that's extremely hot with investors right now,
[20:32] Speaker 12: defense tech broadly. Hadrian makes facilities to speed up domestic
[20:37] Speaker 12: manufacturing and that's been hugely of interest to a lot
[20:40] Speaker 12: of folks and Silicon Valley.
[20:42] Speaker 2: Thank you for that.
[20:43] Speaker 3: I would also add that s fokes first and Hadrium
[20:45] Speaker 3: said the reporting was incorrect but declined to add further context.
[20:49] Speaker 2: I would have done that for you. Thank you very much.
[20:51] Speaker 2: There is some reporting that we did about where they.
[20:54] Speaker 3: Approached debt or a lack of approach to debt.
[20:58] Speaker 12: Just explain that bites so building large automated factories often
[21:03] Speaker 12: means that companies need to take on debt to take
[21:06] Speaker 12: you know, on the risks of building these large facilities.
[21:11] Speaker 12: Hadrian currently has four factories, the latest which operate, which
[21:15] Speaker 12: opened in March off the back of a two point
[21:18] Speaker 12: four billion dollar contract with the US Navy. So this amount,
[21:22] Speaker 12: this up to one billion dollars is just equity that
[21:25] Speaker 12: they would raise. It doesn't include any debt that they
[21:27] Speaker 12: may be discussing.
[21:28] Speaker 3: Rebecca, you did a great job explaining, you know what
[21:31] Speaker 3: Hadrian does. It's basically an amazing factory or contract manufacturer. Generally,
[21:37] Speaker 3: what is bench capital attitude right now to defense technology
[21:40] Speaker 3: and this whole kind of reindustrializing America push.
[21:45] Speaker 2: There's tons of excitement here.
[21:47] Speaker 12: Ed investors in Hadrin include Andres and Horowitz through its
[21:50] Speaker 12: American Dynamism Fund, Pedersfield's founder's fund, Lux Capital. This effort
[21:55] Speaker 12: to sort of reshore manufacturing and create the cycle where
[22:00] Speaker 12: consumers and companies can just rely on production and consumption
[22:05] Speaker 12: within the US is a huge interesting.
[22:09] Speaker 3: These people Bloombergs, Rebecca Torrents teaming up with me on
[22:12] Speaker 3: that one on Hadrian potentially hitting seven point five billion
[22:15] Speaker 3: dollar valuation, Thank you very much. Now, coming up later
[22:18] Speaker 3: in the show, we're going to sit down with David
[22:20] Speaker 3: George and Drees and horror Itz general partner on a
[22:23] Speaker 3: sixteen c's big bet on SpaceX conversation that we're really
[22:27] Speaker 3: looking forward to right now. This is what technology looks
[22:31] Speaker 3: like in financial markets, particularly inequity markets, and we are
[22:35] Speaker 3: all waiting for what is a big one that is
[22:38] Speaker 3: Micron reporting earnings after the closing bell right now, Na's
[22:43] Speaker 3: like one hundred modestly hired, a flat underperformance in chip
[22:46] Speaker 3: stocks and bitcoin just shy of sixty one thousand US
[22:49] Speaker 3: dollars per token. Welcome back to Bloomberg Tech. It's a
[23:01] Speaker 3: big moment in time for Memory. The biggest headline is
[23:04] Speaker 3: probably the sk Heinex will try and raise twenty nine
[23:07] Speaker 3: billion dollars in a US listing.
[23:10] Speaker 2: But the AI trade is.
[23:11] Speaker 3: Going to get its next queue from Micron's earnings after
[23:14] Speaker 3: the bell. There's a lot of literature on the Bloomberg
[23:16] Speaker 3: about that. The stocks down about one zero point three
[23:19] Speaker 3: percent right now. It's highly analogous from what we saw
[23:22] Speaker 3: in some of the earlier days of Nvidia twenty twenty
[23:24] Speaker 3: three to twenty twenty five, massive growth year on year,
[23:28] Speaker 3: both on the top and bottom line. But the market
[23:31] Speaker 3: really wants a bullish signal, not that the bottleneck that
[23:35] Speaker 3: is Memory is unwinding a little on the AI infrastructure
[23:38] Speaker 3: build out, but that the demand has staying power. So
[23:41] Speaker 3: that is what we're watching for throughout the course of
[23:43] Speaker 3: the day. Another top story, Elon Musk is reshaping his
[23:47] Speaker 3: empire's balance sheet. SpaceX just pulled off a record twenty
[23:51] Speaker 3: five billion dollar investment grade bond sale, taking on more
[23:54] Speaker 3: debt while lowering overall borrowing costs across must sprawling businesses.
[23:59] Speaker 3: For more Bloombergs, Private credit reporter Paula Sellison joins us,
[24:03] Speaker 3: how did it go?
[24:04] Speaker 11: It went very well that the peak of demand. There
[24:07] Speaker 11: were eighty nine billion dollars in orders for what was
[24:10] Speaker 11: ultimately a twenty five billion dollar bond deal that allowed
[24:13] Speaker 11: the company to tighten or get lower borrowing costs over
[24:16] Speaker 11: the course of the marketing process, and investors are very
[24:19] Speaker 11: eager to lend.
[24:21] Speaker 2: There is a bit of a history lesson in this.
[24:23] Speaker 2: Find it fascinating.
[24:24] Speaker 3: So, yeah, must takes Twitter private, x joins Xai, Xai
[24:30] Speaker 3: merges with SpaceX, and then SpaceX does the biggest IPO
[24:33] Speaker 3: in history. And if you look at how they kind
[24:35] Speaker 3: of shuffled the deck with the proceeds of the bond
[24:37] Speaker 3: sale and what they planned to do in paying down
[24:40] Speaker 3: existing loans, it kind of takes us back to square one?
[24:43] Speaker 2: Is that right?
[24:44] Speaker 11: Well, this was really all one long, convoluted way of
[24:48] Speaker 11: allowing Xai to become part of an investment grade company
[24:51] Speaker 11: and then be able to borrow at those borrowing costs. So,
[24:54] Speaker 11: you know, we had the Twitter LBO which ended up
[24:57] Speaker 11: adding about thirteen billion dollars of debt, and then the
[24:59] Speaker 11: Xai leverage loan and hiled bond sales which added five
[25:02] Speaker 11: billion dollars of debt, So when SpaceX acquired XAI, there
[25:06] Speaker 11: was seventeen point five billion dollars of debt with interest
[25:09] Speaker 11: rates in the nine and a half percent to twelve
[25:11] Speaker 11: and a half percent range. All of that essentially just
[25:13] Speaker 11: got refinanced initially through a twenty billion dollar bridge loan
[25:17] Speaker 11: that was then taken out with the investment grade bond
[25:19] Speaker 11: sale on Tuesday, and now interest rates around five and
[25:22] Speaker 11: a half to six.
[25:22] Speaker 2: And a half percent.
[25:24] Speaker 3: We broke the story in just the days before the
[25:27] Speaker 3: IPO that SpaceX had got IG rating from the free
[25:29] Speaker 3: main agencies. That was really important, right The story right
[25:33] Speaker 3: now in the AI space is borrowing and the IG
[25:37] Speaker 3: rating has proved to be pretty handy to SpaceX.
[25:41] Speaker 11: Absolutely, so there is so much need for capital. Investment
[25:44] Speaker 11: companies just can't do it with cash flow. They have
[25:46] Speaker 11: to borrow, and they're borrowing across all the different debt markets.
[25:49] Speaker 11: But the best place to borrow for most of them
[25:51] Speaker 11: is the investment grade bond market because it's very large,
[25:54] Speaker 11: very deep pockets, and it's a cheaper borrowing cost. So
[25:57] Speaker 11: being able to get those investment grade ratings just opens
[25:59] Speaker 11: up access to capital that makes it so much easier
[26:02] Speaker 11: to borrow all this money to finance the AI build
[26:04] Speaker 11: out the most.
[26:06] Speaker 2: Paula Cellison, thank you very much.
[26:08] Speaker 3: Indeed, in the equity space SpaceX actually hire again today
[26:10] Speaker 3: one hundred and fifty eight dollars per share SpaceX David,
[26:16] Speaker 3: George Andresen horror it's general partner, is on set with
[26:19] Speaker 3: us right now. He led the firm's investment investment in SpaceX,
[26:21] Speaker 3: a state which is now valued at more than ten
[26:23] Speaker 3: billion dollars and it's proved to be Andresen's largest return
[26:27] Speaker 3: in history. And you're somebody that I've wanted to speak
[26:30] Speaker 3: to about the company for a long time. There's a
[26:32] Speaker 3: lot here, but I think just as an opening reflect
[26:35] Speaker 3: on the IPO, David, and I guess not just in
[26:39] Speaker 3: a moment in time for Andresen, but what you think
[26:42] Speaker 3: it's signified for what is a very big and now
[26:45] Speaker 3: quite diverse company.
[26:47] Speaker 13: Yeah, well, look, thanks for having me on. A great
[26:48] Speaker 13: to be here with you. The IPO is just a
[26:51] Speaker 13: milestone in the history of the company. I think it's
[26:55] Speaker 13: a great event for the company to be able to
[26:59] Speaker 13: access a new investor's access capital that they couldn't otherwise access.
[27:03] Speaker 2: Before.
[27:04] Speaker 13: And one of the things that I'm most excited about
[27:06] Speaker 13: as relates to the IPO is the fact that retail
[27:09] Speaker 13: was able to partake in such a big fashion. So
[27:12] Speaker 13: if you look at the allocation that went to retail
[27:15] Speaker 13: as a proportion of the overall IPO, it's around thirty percent.
[27:19] Speaker 13: And so if you just take that portion that went
[27:21] Speaker 13: to retail, it's almost the largest IPO by itself in
[27:25] Speaker 13: the history of IPOs.
[27:26] Speaker 2: So I think it's fantastic that. Why is that part
[27:28] Speaker 2: of that? David, Sorry to interrupt you.
[27:30] Speaker 13: Retail has proven to be an excellent group of buyers
[27:33] Speaker 13: and holders of stock. So obviously it's a large proportion
[27:37] Speaker 13: of the ownership of Tesla, but they also are large
[27:40] Speaker 13: owners of the meg seven at large, and so I
[27:43] Speaker 13: think it's excellent for SpaceX to be able to access that.
[27:46] Speaker 13: I think you know, in large part, retail is long
[27:48] Speaker 13: term thinking, which aligns with the way Elon thinks and
[27:51] Speaker 13: how he's built SpaceX, and so I love that they're
[27:53] Speaker 13: a big part of it.
[27:54] Speaker 3: Probably the question I get for you most commonly is
[27:57] Speaker 3: about the lock up.
[27:58] Speaker 2: What will happen?
[27:59] Speaker 3: You know, to the venture firms would say, well, we
[28:02] Speaker 3: will redistribute to our LPs the stock and we expect
[28:05] Speaker 3: lots of them to hold. But that would seem to
[28:07] Speaker 3: make the retail holders important. And what was like a
[28:11] Speaker 3: very complex lockup, It wasn't straightforward.
[28:13] Speaker 13: Yeah, look, I think this is a great step forward
[28:16] Speaker 13: generally and how lockups are run, and I expect for
[28:19] Speaker 13: IPOs going forward, this is probably going to be the
[28:21] Speaker 13: way it's run. So cerebristed something similar where you have
[28:25] Speaker 13: the lockups come off gradually over time, and I think
[28:28] Speaker 13: this is good, this is healthy.
[28:30] Speaker 2: It allows you know.
[28:32] Speaker 13: The public market investors to be able to buy stock
[28:34] Speaker 13: over time as opposed to you know at cliff events.
[28:37] Speaker 13: For us, we're very long term thinkers. You know, we're
[28:39] Speaker 13: happy shareholders of SpaceX. You know, we think very much
[28:43] Speaker 13: about the long term.
[28:45] Speaker 12: You know.
[28:45] Speaker 13: Obviously, what makes us so excited about the business is
[28:49] Speaker 13: all of the things that can go right for SpaceX,
[28:51] Speaker 13: all the optionality that you have. And so the way
[28:54] Speaker 13: that we look at it is you start with the
[28:55] Speaker 13: launch business, and in order to be successful as a
[28:58] Speaker 13: space business, you have to rely be able to get
[29:00] Speaker 13: things to space. And so SpaceX has the infrastructure in
[29:05] Speaker 13: an incredible way to be able to build applications on top. Obviously,
[29:08] Speaker 13: Starlink is the first one. We expect that with Starship
[29:12] Speaker 13: and V three, they've achieved some de risking in terms
[29:16] Speaker 13: of what they can bring up to space. Eventually they'll
[29:19] Speaker 13: have rapid reusability with multiple launches per day on Starship.
[29:24] Speaker 13: It's a feat of physics magnificence that they can take
[29:29] Speaker 13: something that's the size of larger than a football field,
[29:32] Speaker 13: send it up to space, grab it with chopsticks, and
[29:35] Speaker 13: then reuse it again. They'll rapidly be able to do that,
[29:38] Speaker 13: and that's going to enable all the applications that they
[29:40] Speaker 13: want to build on top.
[29:41] Speaker 3: It also now has a name starmind Yes as of
[29:45] Speaker 3: last night, and you know Elon Musk engaging with others
[29:48] Speaker 3: on X. There are many out there that believe that
[29:52] Speaker 3: the timeline that was presented in the perspectus of Orbital
[29:55] Speaker 3: Date Center SPACEXIT as early as twenty twenty eight, actually
[29:59] Speaker 3: we could see some pool forward on that. Again linked
[30:01] Speaker 3: to the tracking of Starship. Where do you sit in
[30:06] Speaker 3: what's realistic.
[30:07] Speaker 13: Near to Yeah, Look, I think the biggest hurdle to
[30:10] Speaker 13: cross is just rapid reusability of Starship, and I think
[30:15] Speaker 13: you know, there's a consistent theme with Elon's companies, and
[30:17] Speaker 13: certainly in SpaceX, which is physics has been de risked,
[30:21] Speaker 13: and so now.
[30:22] Speaker 2: It becomes an execution question.
[30:23] Speaker 13: And so as you think about timelines, we feel like
[30:26] Speaker 13: it is inevitable that they can achieve these milestones, just
[30:29] Speaker 13: a question of how quickly and at what cost. And
[30:32] Speaker 13: so if you have starship that can rapidly, you know,
[30:35] Speaker 13: reliably get things up and back to space, we think
[30:38] Speaker 13: that's the major unlock for bringing compute to space.
[30:42] Speaker 2: I know earlier you.
[30:43] Speaker 13: Were talking about the eight hundred and fifty billion dollars
[30:45] Speaker 13: of CAPEX for AI data centers on terrestrial grounds. Right
[30:51] Speaker 13: now it's getting harder and harder to get data centers
[30:54] Speaker 13: live on the ground here. I think it's a matter
[30:58] Speaker 13: of time before we have it in space. I actually
[31:01] Speaker 13: think I like to reframe it a little bit. I
[31:03] Speaker 13: don't talk about it as orbital data centers. I talk
[31:06] Speaker 13: about them as sort of airplane sized.
[31:09] Speaker 2: Gpu ras in space.
[31:10] Speaker 13: So think of it as you know, something like seventy
[31:13] Speaker 13: two GPUs up in space with big wings that are
[31:17] Speaker 13: solar arrays that are kind of the size of a
[31:19] Speaker 13: seven thirty seven. And then you can have many, many
[31:21] Speaker 13: of those in space. They've demonstrated that they can do this.
[31:24] Speaker 13: They have ten thousand LEO satellites, So we feel like
[31:27] Speaker 13: it's a matter of time and execution before they can
[31:30] Speaker 13: do that. The physics has been de risked, you know,
[31:33] Speaker 13: to the point about how it's getting harder and harder.
[31:35] Speaker 2: To do this.
[31:36] Speaker 13: You know, on the ground, I think at a minimum,
[31:39] Speaker 13: orbital data centers will be incremental capacity that you can
[31:43] Speaker 13: have in space on top of what we have on Earth.
[31:46] Speaker 13: And I think there's a case that in the fullness
[31:48] Speaker 13: of time, the economics actually get better than on the ground.
[31:52] Speaker 3: David the firm Andreessen Horowitz has been involved with must
[31:56] Speaker 3: companies in different ways for about six years, in different
[31:59] Speaker 3: ways x XAI, the different financial transactions that took place.
[32:04] Speaker 2: For you, how much was this this.
[32:06] Speaker 3: IPO a referendum essentially on Musk himself or how central
[32:12] Speaker 3: to the fate of this company do you see long being?
[32:16] Speaker 13: Yeah, Look, Elon is the centerpiece of all of those
[32:19] Speaker 13: investment thesis that we had right in backing his companies.
[32:23] Speaker 13: You know, he's he's been remarkably strategic and how he's
[32:26] Speaker 13: put the companies together, and you know, he has done
[32:29] Speaker 13: very well by shareholders.
[32:30] Speaker 2: He's taking care of shareholders.
[32:31] Speaker 13: He takes a lot of pride in that, and we
[32:33] Speaker 13: appreciate that As his partner. I think as it relates to,
[32:37] Speaker 13: you know, the go forward, he's been very smart. He
[32:39] Speaker 13: and Brett Johnson and the team ASPACEX have been very
[32:42] Speaker 13: very smart about capital allocation, right, you know, when they
[32:45] Speaker 13: have done acquisitions, they have been remarkably strategic in terms
[32:49] Speaker 13: of putting things together first, you know, and putting X
[32:52] Speaker 13: with Xai, which is very logical, and then obviously the
[32:55] Speaker 13: fit of XAI and SpaceX together, you know, is undeniable.
[32:58] Speaker 13: So we expect that will continue to be very smart
[33:01] Speaker 13: about capital allocation, take care of shoreholders, and we appreciate that.
[33:04] Speaker 3: Where do you stand, David then, on the transaction that
[33:08] Speaker 3: everyone continues to talk about, which is a future where
[33:10] Speaker 3: Tesla merges with SpaceX, Is it rational to your mind?
[33:14] Speaker 13: I go back to the things that he's done in
[33:17] Speaker 13: the past, and the things that he's done in the past,
[33:19] Speaker 13: is he has decided to do acquisitions of his companies
[33:23] Speaker 13: or mergers of his companies when there is very strong
[33:26] Speaker 13: strategic alignment and it makes business sense.
[33:28] Speaker 2: So I wouldn't expect it to be any different going forward.
[33:32] Speaker 3: There's this broader idea, if you extract it out from
[33:35] Speaker 3: from Elon Musk, that even at the growth stage, late
[33:39] Speaker 3: private companies staying private for longer, late stage growth, that
[33:42] Speaker 3: the investment thesis is still around the founder, the person
[33:46] Speaker 3: at the top. I don't know, how whether you share
[33:48] Speaker 3: that view of others that even if you are talking
[33:51] Speaker 3: in the tens hundreds of billions of dollars of value,
[33:55] Speaker 3: that value is assignable to an individual or a group
[33:58] Speaker 3: of individuals.
[33:59] Speaker 2: Yeah.
[33:59] Speaker 13: Look, I think you can see reflections of this actually
[34:02] Speaker 13: in the public markets. So I think you can see
[34:04] Speaker 13: this in Elon companies. Certainly. I think you could see this,
[34:07] Speaker 13: you know, with Apple under Steve Jobs. I think you
[34:11] Speaker 13: could see it with Meta under Mark Zuckerberg. But certainly
[34:14] Speaker 13: in the private markets. You know, this is a centerpiece
[34:18] Speaker 13: of our investment thesis when we're backing companies at any
[34:21] Speaker 13: stage and any size. I recently wrote a piece that
[34:25] Speaker 13: basically said late stage venture, which is our asset class,
[34:29] Speaker 13: is not about capital markets.
[34:30] Speaker 2: It's about founders.
[34:31] Speaker 13: Late stage venture is about late stage founders and enabling
[34:34] Speaker 13: them to think long term, make big decisions, and make
[34:38] Speaker 13: big bets, and oftentimes they choose to do that in
[34:41] Speaker 13: the private markets.
[34:42] Speaker 2: And so you know, our asset class.
[34:44] Speaker 13: That we play in, it's about five trillion dollars in size.
[34:47] Speaker 13: It's grown ten x over the last ten years. It's
[34:50] Speaker 13: larger than the Russell two thousand. It's almost twenty percent
[34:52] Speaker 13: as large as the Nasdaq. So this has become a
[34:55] Speaker 13: real asset class. And I think there's reasons that we
[34:58] Speaker 13: should explore. Why is it that cost a title?
[35:00] Speaker 2: What would you call that late stage venture?
[35:02] Speaker 13: Late stage privates with the value crews, where the value yeah,
[35:06] Speaker 13: and where the value crews, yeah. And the centerpiece of
[35:09] Speaker 13: the investment thesis is around the founder and the big
[35:11] Speaker 13: decisions that the founder has to make.
[35:13] Speaker 2: It's a lot of key man risk to model for.
[35:15] Speaker 13: Yes, that's right, But I think the more important thing
[35:17] Speaker 13: to model for is what can go right? Okay, you
[35:19] Speaker 13: know we always ask ourselves, you know, it's hard to
[35:23] Speaker 13: imagine what could go right. And we even see this
[35:25] Speaker 13: with mature companies. If you look back at the time
[35:28] Speaker 13: when you know the iPhone came out, If you look
[35:30] Speaker 13: at the two thousand and nine consensus analyst estimates for
[35:34] Speaker 13: what Apple was going to do, and that in two
[35:37] Speaker 13: thousand and nine. For the next four years, and then
[35:39] Speaker 13: fast forward four years, they actually beat those numbers by
[35:41] Speaker 13: three x. And that's probably the most covered company in
[35:44] Speaker 13: the world. So we like to think with a great founder,
[35:46] Speaker 13: with a technology trend on its.
[35:48] Speaker 2: Back, what can go right?
[35:49] Speaker 13: And you know, if you look at this last cycle
[35:52] Speaker 13: that we've just gone through pre AI, you know, this
[35:55] Speaker 13: cycle of mobile phones, social e commerce, SaaS, cloud all
[36:00] Speaker 13: put together, the big story is that that big trend
[36:03] Speaker 13: added about twenty five to thirty.
[36:05] Speaker 2: Trillion of new market cap.
[36:07] Speaker 13: And I happen to think that on this AI wave,
[36:09] Speaker 13: on the back of this AI wave, the.
[36:11] Speaker 2: Trend's going to be even bigger.
[36:13] Speaker 3: You head a growth team that manages twenty two billion
[36:16] Speaker 3: dollars across five funds. But if you think about it holistically,
[36:20] Speaker 3: you're basically in the top fifteen private companies by valuation,
[36:25] Speaker 3: which would have included SpaceX before it went public. How
[36:29] Speaker 3: are you now deploying capital new capital into new companies,
[36:33] Speaker 3: you know, with those existing investments in mind.
[36:36] Speaker 13: Yeah, Look, our whole thesis is we want to be
[36:39] Speaker 13: involved in the best companies at the earliest stage possible
[36:42] Speaker 13: and then continue to back them every step of the
[36:44] Speaker 13: way as they need capital, and often that has been
[36:48] Speaker 13: through late stage rounds in the.
[36:49] Speaker 3: Private market because you have a newer fund, right which
[36:52] Speaker 3: you can deploy capital out there.
[36:54] Speaker 2: We do.
[36:54] Speaker 13: Yeah, we we have our LSV five Late Stage Venture
[36:58] Speaker 13: Fund five as we call it, it's our fifth fund.
[37:00] Speaker 13: It's about a seven billion dollar fund that we're deploying
[37:03] Speaker 13: out of currently, and we see a ton of really
[37:06] Speaker 13: interesting opportunities. So you know, if you go back to
[37:08] Speaker 13: the market that we're seeing and just look at our
[37:12] Speaker 13: portfolio at a sixteen Z and our latest fund, our
[37:15] Speaker 13: portfolio dollar weighted is growing over one hundred percent year
[37:18] Speaker 13: over year. If you compare that to the size similar
[37:21] Speaker 13: sized companies in the public markets, those companies are growing
[37:24] Speaker 13: about twenty percent year over year. So we put a
[37:26] Speaker 13: huge premium on very fast growth, great founders, market leadership,
[37:30] Speaker 13: and sort of building.
[37:31] Speaker 2: On the back of the big trends.
[37:33] Speaker 3: Also noting the cap tables look a bit different like
[37:35] Speaker 3: with the mutual funds private growth equity. David George Havenjreson Horowitz,
[37:40] Speaker 3: head of Growth and general partner.
[37:42] Speaker 2: Really grateful to have you on Bloomberg Tech. Thank you
[37:43] Speaker 2: very much.
[37:44] Speaker 3: Coming up, Cerebris reported its first ever quarterly earnings since
[37:48] Speaker 3: going public. CEO Andrew Feldman's joining us. The stock is
[37:51] Speaker 3: down seventeen percent. There's a lot to unpack there. This
[37:54] Speaker 3: is Bloomberg Tech. Crebress reported quarterly earnings for the first
[38:12] Speaker 3: time since going public last month. It's sales outlook beat
[38:15] Speaker 3: Wall Street estimates, but still disappointed investors hoping to see
[38:19] Speaker 3: the company carve out a bigger slice of the aident
[38:22] Speaker 3: center market. Right now, shares down seventeen point five percent,
[38:26] Speaker 3: its biggest drop in its quite short history as a
[38:29] Speaker 3: public company. What's behind that? CEO Andrew Feldman's with us
[38:32] Speaker 3: and welcome back to Bloomberg Tech.
[38:34] Speaker 2: Andrew.
[38:34] Speaker 3: You know, there was a time where we would talk
[38:36] Speaker 3: about the merits of top to tail server ownership, how
[38:40] Speaker 3: owning all of the content of Now we're going to
[38:41] Speaker 3: talk about margin contraction, and we're going to talk about
[38:44] Speaker 3: the stock being down seventeen percent. That to me is
[38:48] Speaker 3: kind of the mismatch that the outlook on the sales
[38:50] Speaker 3: side beat Wall Street estimates. I think a lot of
[38:53] Speaker 3: people are trying to understand the sequential margin decline and
[38:57] Speaker 3: for me that this is about ramping output for two
[39:00] Speaker 3: big customers Is that true.
[39:02] Speaker 5: Yeah.
[39:03] Speaker 14: I think what we did is we put forward a
[39:05] Speaker 14: plan in the start of twenty six, we shared it
[39:09] Speaker 14: with investors as we went public, and we're ahead of plan.
[39:13] Speaker 5: You know. We delivered record.
[39:14] Speaker 14: Revenues of one hundred and ninety one million of ninety
[39:17] Speaker 14: two percent year over year, and for our cloud business,
[39:21] Speaker 14: you know, it was up one hundred and sixty seven
[39:23] Speaker 14: percent year over year. We beat margin consensus substantially, and
[39:30] Speaker 14: then we guided for full year the growth.
[39:33] Speaker 5: Margins would be ten percent better than plant.
[39:36] Speaker 14: We also shared the in Q two and Q three
[39:40] Speaker 14: we would go back to some of our customers and
[39:44] Speaker 14: we would rent back year that we'd sold them to
[39:46] Speaker 14: try and keep up with demand, and then that would
[39:48] Speaker 14: have a margin impact on the order ten.
[39:50] Speaker 5: Or fifteen points.
[39:52] Speaker 14: We did that to keep our customers close to be
[39:56] Speaker 14: sure we could keep up with their extraordinary demand for
[40:00] Speaker 14: or for our product for fast inference, and so that
[40:05] Speaker 14: that was the story.
[40:09] Speaker 5: On every metric we put out, we're ahead of plan.
[40:13] Speaker 2: Have the proceeds from.
[40:14] Speaker 3: The IPO actually allowed you to move more quickly in
[40:17] Speaker 3: ramping up capacity?
[40:19] Speaker 14: Yeah, I think capacity is the largest constraint right now
[40:24] Speaker 14: for everyone.
[40:25] Speaker 5: Data centers are and.
[40:27] Speaker 14: We've significantly increased our our ability and our pipeline for
[40:34] Speaker 14: data centers, which is now very large. You know, we
[40:40] Speaker 14: announced a data center partnership with Bell Canada for one
[40:44] Speaker 14: hundred and twenty megawatts that will be delivered in twenty
[40:47] Speaker 14: twenty seven. We are pursuing data centers across the US,
[40:53] Speaker 14: in Canada, in Europe, in the Middle East.
[40:56] Speaker 5: There are.
[40:59] Speaker 14: The resources that we have now at our disposal give
[41:04] Speaker 14: us tremendous advantage in the pursuit of this The limiting
[41:07] Speaker 14: factor data centers.
[41:10] Speaker 3: What you're talking about, like matter of factly, is buildings
[41:14] Speaker 3: not necessarily the compute, right It's not what you guys
[41:17] Speaker 3: are offering. How difficult right now is it is to
[41:20] Speaker 3: get moving in America or other markets, to get planning approval,
[41:24] Speaker 3: get the concrete, get the labor, get the thing built.
[41:27] Speaker 14: Now that that's the irony of this market, that the
[41:32] Speaker 14: AI market is moving at blistering speed and we are
[41:36] Speaker 14: being constrained by data centers which move with the speed
[41:39] Speaker 14: of real estate. And so that is a problem that
[41:43] Speaker 14: is being confronted by by everybody in the category, by
[41:48] Speaker 14: the hyperscalers, by the neo clouds, by the new generation clouds.
[41:53] Speaker 14: Everybody is confronting this similar problem.
[41:56] Speaker 3: Andrew Cerebris does not rely on traditional chip HBM. Would
[42:01] Speaker 3: you just explain that the basics of the technology, But
[42:05] Speaker 3: how insulated are you from the memory bottleneck that others
[42:08] Speaker 3: are experiencing.
[42:09] Speaker 5: Yeah, that's a really good point.
[42:12] Speaker 14: Because of our innovative architecture, because of our wafer scale approach,
[42:16] Speaker 14: we don't use HBM. HBM is a type of DRAM
[42:21] Speaker 14: and it's made by three companies, one of whom is
[42:23] Speaker 14: reporting shortly right. That's Microns, Heenex and Samsung. There's a
[42:28] Speaker 14: global shortage, it's extremely expensive, lead times are long, and
[42:32] Speaker 14: we don't use it, so we have a tremendous advantage there.
[42:36] Speaker 14: The other constraints in the supply chain for many are cooths,
[42:42] Speaker 14: which is a process inside of TSMC.
[42:44] Speaker 5: Again we don't use it.
[42:46] Speaker 14: And the third is a capacity at the three nanometer node.
[42:51] Speaker 14: That's space in TSMC's factory that makes three and nanimeter chips.
[42:56] Speaker 5: Again we don't use it. We're at the five nanometer node.
[43:00] Speaker 14: Our architecture has allowed us to deliver the fastest inference
[43:03] Speaker 14: in the world by an order of magnitude while avoiding
[43:08] Speaker 14: the main supply supply chain constraints faced by others in
[43:13] Speaker 14: the field.
[43:15] Speaker 3: Can you say, hand on heart, not just winning business,
[43:18] Speaker 3: but if you actually been able to go to a
[43:20] Speaker 3: customer and say we can get this compute online faster
[43:23] Speaker 3: than others for those reasons you just outlined and then actually.
[43:26] Speaker 2: Gone and done it.
[43:27] Speaker 5: Oh for sure.
[43:29] Speaker 2: I mean any cases I.
[43:30] Speaker 14: Signed, we signed. For example, we signed our contract with
[43:36] Speaker 14: open Ai on December twenty fourth and had them in
[43:42] Speaker 14: for production on February first. That's unheard of.
[43:46] Speaker 3: So that's quick as I went really quickly Andrew this morning.
[43:50] Speaker 3: Open Ai is out with Kalapino Intelligence processor. But everyone
[43:55] Speaker 3: is compute constrained, right, How do you interpret the Frontier
[43:59] Speaker 3: Labs going to custom and a six alongside their other
[44:02] Speaker 3: compute options.
[44:04] Speaker 12: Ed.
[44:04] Speaker 14: One of the things we've been saying all along is
[44:06] Speaker 14: that this market is enormous and is going to be
[44:10] Speaker 14: met with a heterogeneous collection of architectures for hardware. This
[44:15] Speaker 14: market will not be consolidated around GPUs. There will be
[44:19] Speaker 14: a six, There will be a six from Hyperscalers, there
[44:22] Speaker 14: will be a six from from Labs, and then there'll
[44:26] Speaker 14: be companies like Cerebrase with pioneering architectures who all of
[44:31] Speaker 14: us will will will take big bites of this enormous market.
[44:35] Speaker 14: I think one of the things that's difficult to get
[44:37] Speaker 14: your head around is just how big the compute market
[44:41] Speaker 14: is right now. That one of the things AI does
[44:44] Speaker 14: is it makes tractable for compute much of the world
[44:48] Speaker 14: around us, and that really wasn't the case prior to AI.
[44:53] Speaker 3: Andrew I it's if the world is simple, and it's
[44:55] Speaker 3: money that greases the wheels to get your industry going.
[44:58] Speaker 3: Can we see you come back to capital markets in
[45:01] Speaker 3: some form just so you can move quicker.
[45:04] Speaker 14: We have more than nine billion on the balance sheet,
[45:06] Speaker 14: I think, so we're really pleased with our position there.
[45:12] Speaker 14: But of course we are always scanning both the capital markets,
[45:16] Speaker 14: both equity and debt for ways to accelerate our growth.
[45:22] Speaker 3: Finally, how are you judging success yourself? What is the
[45:25] Speaker 3: milestone that you'd pitch to the market to keep closest
[45:28] Speaker 3: attention to.
[45:30] Speaker 14: Well, I think I think if my mother's proud of me,
[45:33] Speaker 14: I think that's the biggest.
[45:35] Speaker 5: Thing you can you can ask for.
[45:38] Speaker 14: I think for markets, I think when you lay out
[45:43] Speaker 14: a plan that is aggressive and you crush it, Uh,
[45:48] Speaker 14: that's how you feel good about both your ability to
[45:51] Speaker 14: execute in your ability to predict your own execution.
[45:54] Speaker 5: And so you know, we announced yesterday that.
[45:58] Speaker 14: That we would beat our our full year we give
[46:01] Speaker 14: full year guidance that was ten gross margin points above concernsus.
[46:07] Speaker 14: You ought to be proud of that, and we got
[46:09] Speaker 14: to continue to execute and continue to set extraordinarily high
[46:12] Speaker 14: barrows and then continue to.
[46:16] Speaker 5: Beat them.
[46:18] Speaker 3: Cerebra Ceo, Andrew Feldman, thank you very much. Indeed, really
[46:23] Speaker 3: Andrew alluded to it. Micron is the big one after
[46:25] Speaker 3: the closing bell. Top line growth almost three hundred percent
[46:28] Speaker 3: year on year, net income growth twelve hundred percent.
[46:31] Speaker 2: Year on year. The bar is very very high the market.
[46:35] Speaker 3: It's very analogous to Nvidia what we saw twenty twenty
[46:38] Speaker 3: two to twenty twenty five, massive growth, but investors always
[46:42] Speaker 3: wanting more. The stocks down eight ten percent going into
[46:45] Speaker 3: that print. Another stock we're watching today US fast food
[46:49] Speaker 3: chain Wendy's surging as much as forty two percent.
[46:52] Speaker 2: But wait, what does that have to do with tech?
[46:55] Speaker 3: Well, it appears Wendy's has become the latest memes stock
[46:59] Speaker 3: target shot up the rankings on stock Twitz climb to
[47:02] Speaker 3: the top of the platform's trending list. Seems to have
[47:04] Speaker 3: been sparked by a now deleted post on Reddit's Wall
[47:07] Speaker 3: Street Bets forum that urged members to save Wendy's before
[47:10] Speaker 3: it's too late. That does it for this edition of
[47:12] Speaker 3: Bloomberg Tech. Recap everything on the podcast. You know where
[47:15] Speaker 3: to find it online, Apple, Spotify and iHeart and all
[47:18] Speaker 3: the Bloomberg platforms.
[47:19] Speaker 2: This is Bloomberg Tech.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.
