# Nvidia Makes $3.5 Billion Bet on MediaTek

An episode of [Bloomberg Tech](https://invisiblenetwork.ai/podcast/1611767434), 31 Aug 2026, 43 min.

Audio: https://podtrac.com/pts/redirect.mp3/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/41764a4f-fc64-4e11-89ba-ae7c0030ab5e/adc72848-2317-4f87-8e97-b4b70117204b/audio.mp3?utm_source=Podcast&in_playlist=9caafc41-289c-4115-995d-ae7c0030ab75

Bloomberg’s Ed Ludlow breaks down Nvidia's $3.5 billion investment in MediaTek, deepening its collaboration with the Taiwanese chipmaker. Plus, today is Tim Cook's last day as CEO of Apple before John Ternus takes the reins tomorrow, and all eyes are on Anthropic as it prepares to file publicly in the coming weeks. See omnystudio.com/listener for privacy information.

## Transcript

[0:02] Speaker 1: Bloomberg Audio Studios.
[0:04] Speaker 2: Podcasts.
[0:05] Speaker 1: Radio. News.
[0:13] Speaker 3: Bloomberg Tech is live from the heart of Silicon Valley
[0:17] Speaker 3: with Ed Ludlow in San Francisco.
[0:23] Speaker 2: This is Bloomberg Tech.
[0:24] Speaker 4: Coming up, NVIDIA makes a $ 3. 5 billion investment in MediaTek,
[0:28] Speaker 4: deepening its collaboration with the Taiwanese chipmaker.
[0:31] Speaker 2: We'll hear from both CEOs.
[0:33] Speaker 4: Plus, today is Tim Cook's last day as CEO of
[0:37] Speaker 4: Apple before John Ternus takes the reins tomorrow. And all
[0:41] Speaker 4: eyes on Anthropic as it prepares to file publicly in
[0:43] Speaker 4: the coming weeks. This is other companies either own U.S.
[0:46] Speaker 4: listing plans in the second half of the year.
[0:50] Speaker 2: Okay, let's get straight to our top story.
[0:52] Speaker 4: NVIDIA's making another multi-billion dollar bet on the AI ecosystem,
[0:56] Speaker 4: investing $ 3. 5 billion in Taiwan's MediaTek. The partnership brings MediaTek's
[1:02] Speaker 4: custom AI chips into NVIDIA's architecture, plugging so-called XPUs into
[1:07] Speaker 4: NVIDIA high-speed networking and memory technology.
[1:10] Speaker 2: The deal could also.
[1:11] Speaker 4: Revive concerns about circular financing as NVIDIA invests across its
[1:16] Speaker 4: own ecosystem. I sat down with NVIDIA CEO Jensen Wang
[1:19] Speaker 4: and MediaTek CEO Rick Tsai earlier today.
[1:22] Speaker 5: This is not circular because obviously they do their own
[1:26] Speaker 5: business and we do our own business. And MediaTek's already
[1:31] Speaker 5: incredibly profitable, incredibly successful. They are the world's best SOC maker.
[1:38] Speaker 5: They make more SOCs than any company in the world.
[1:42] Speaker 5: It's a great pleasure for us to invest in a
[1:44] Speaker 5: technology company in Taiwan.
[1:45] Speaker 1: It's a great pleasure for us to invest in MediaTek.
[1:48] Speaker 5: This is going to turn out, of course, the investment
[1:51] Speaker 5: is grounded on a deep partnership that's going to be
[1:54] Speaker 5: multiple years. This is a part, the roadmap that we're
[1:58] Speaker 5: executing on right now is a, it's not one or
[2:02] Speaker 5: two years, it's a decade long. And so this partnership
[2:06] Speaker 5: is extremely important to us, and we want to make
[2:08] Speaker 5: sure that we demonstrate every opportunity, take every opportunity to
[2:12] Speaker 5: demonstrate the importance of it, the seriousness by which we
[2:16] Speaker 5: take it, the confidence we have in the partnership, and
[2:18] Speaker 5: the investment that we're making here demonstrates that. But ultimately,
[2:23] Speaker 5: this is going to be a phenomenal investment for us,
[2:26] Speaker 5: and it's going to generate incredible returns.
[2:28] Speaker 1: I have every confidence.
[2:29] Speaker 4: Rick, in July, MediaTek said that it would use its
[2:33] Speaker 4: bond program, $ 5 billion of dollar bonds, basically to support
[2:37] Speaker 4: and grow supply chain. Earlier in our conversation, you talked
[2:41] Speaker 4: about the benefit of working with NVIDIA on MVLink Fusion
[2:44] Speaker 4: is speed to market. What do you plan to do
[2:48] Speaker 4: with the proceeds? And when you talk about supply chain,
[2:51] Speaker 4: do you need to do more on Taiwanese suppliers or
[2:54] Speaker 4: US suppliers in the context of this relationship?
[2:58] Speaker 2: I sent the, uh.
[3:00] Speaker 6: Suddenly, well, back in the end of July, we were
[3:06] Speaker 6: not able to discuss the strategic investment part of the bond. So,
[3:14] Speaker 6: but the proceeds, I think, will be for multiple purposes,
[3:18] Speaker 6: supply chain definitely being one of them, and the supply
[3:22] Speaker 6: chain is a global, I mean, wherever needed to enhance
[3:27] Speaker 6: and enable the customers to But on the other hand,
[3:31] Speaker 6: I think, as Jensen said so eloquently, the investment that
[3:38] Speaker 6: Miratech is a tech company, it's a very strong tech company.
[3:42] Speaker 2: And we will definitely invest if needed. if anything, more into.
[3:49] Speaker 6: The technology in our capability to design, in our capability
[3:53] Speaker 6: to make the ASIC chip in the IP for interconnects
[3:58] Speaker 6: in advanced packaging.
[4:00] Speaker 2: I think the.
[4:02] Speaker 6: The key here is really the collaboration agreements that we
[4:07] Speaker 6: have built and expanded with NVIDIA as described in our
[4:12] Speaker 6: press release. Those are the key things and the proceeds
[4:16] Speaker 6: will basically go in to make those things happen better
[4:21] Speaker 6: and further.
[4:22] Speaker 4: Jenson, on the earnings call last week, you talked about,
[4:25] Speaker 4: with confidence, right, that the GPU-based architecture has superior economics.
[4:31] Speaker 4: You were talking about that in the context of the
[4:33] Speaker 4: Frontier Labs, largely because you were asked about custom silicon.
[4:37] Speaker 4: You know, the simple terms of it is that MVLink
[4:40] Speaker 4: Fusion just makes it easy to take XPU and just
[4:43] Speaker 4: drop it into the data center. That would show that,
[4:47] Speaker 4: you know, Nvidia does see XPU coming into the market.
[4:52] Speaker 4: Just the difference of your thinking this week to what
[4:55] Speaker 4: you outlined last week.
[4:57] Speaker 5: I see XPUs in the market. It's not coming into
[5:01] Speaker 5: the market. It's in the market. Look, an XPU is
[5:04] Speaker 5: a specialized chip. NVIDIA's GPU is a general-purpose accelerator. We
[5:12] Speaker 5: accelerate the entire lifecycle of AI, from data processing to
[5:19] Speaker 5: pre-training to post-training to agentic AI inferencing. We accelerate every
[5:26] Speaker 5: single AI model in the world from closed models, every
[5:29] Speaker 5: closed model, every open model, video models and language models,
[5:34] Speaker 5: biological models and physics models and robotics models. We literally
[5:39] Speaker 5: accelerate every single area of AI. That explains why it
[5:45] Speaker 5: is that we are the most fungible and the most
[5:48] Speaker 5: durable and therefore the most rentable GPU, the most rentable
[5:52] Speaker 5: computer infrastructure in the world.
[5:54] Speaker 1: We're in every cloud, we're on-prem, we're at the edge.
[5:57] Speaker 5: And so we're able to address markets that, frankly, nobody
[6:01] Speaker 5: else can because of the architecture of our company and
[6:05] Speaker 5: the entire full-stack AI factory platform that we have.
[6:12] Speaker 4: Let's get more on the story with Bloomberg's Ian King,
[6:15] Speaker 4: who helped me break it this morning. It's really interesting.
[6:18] Speaker 4: For some time now, gentlemen have been asked about, what
[6:20] Speaker 4: about all these custom chips that hyperscalers are working on,
[6:24] Speaker 4: that the frontier labs are working on? And to this point,
[6:27] Speaker 4: he's been a little bit dismissive of the threat. Today,
[6:30] Speaker 4: with MediaTek, they found a mechanism for those same AI
[6:35] Speaker 4: companies to drop an XPU or a custom chip into
[6:38] Speaker 4: an existing data center to work with NVIDIA's technology.
[6:42] Speaker 2: What else do we need to know here?
[6:43] Speaker 7: Yeah, I mean, taking a step back and looking at
[6:45] Speaker 7: this from a technology perspective, the way the data center
[6:49] Speaker 7: works and how effective it is isn't just how good
[6:52] Speaker 7: are the individual components, it's how good do they all
[6:54] Speaker 7: link together. And what NVIDIA is saying is, hey, use
[6:57] Speaker 7: our link, use this fabric, they call it, and to
[7:01] Speaker 7: bind everything together and it'll all work better. So what
[7:04] Speaker 7: he's trying to do is Even if you want to
[7:06] Speaker 7: go it alone, even if you want to use your
[7:08] Speaker 7: own stuff, use.
[7:09] Speaker 1: It inside our framework.
[7:11] Speaker 7: Use it inside our house and that'll help you.
[7:13] Speaker 1: And of course that means.
[7:15] Speaker 7: in terms of business, that look, hey, if companies want
[7:17] Speaker 7: to do their own thing like Amazon did with that
[7:19] Speaker 7: announcement we saw during earnings, they're going to do it
[7:22] Speaker 7: and they're still going to be using at least some
[7:25] Speaker 7: of his chips, so it works business-wise.
[7:27] Speaker 4: We haven't said it to this point, so I'll say
[7:29] Speaker 4: it now for the audience. This is NVIDIA buying convertible
[7:32] Speaker 4: bonds that convert into equity.
[7:34] Speaker 2: $ 3.
[7:34] Speaker 4: 5 billion, we confirmed, is NVIDIA's biggest non-US direct investment.
[7:40] Speaker 4: So kind of interesting, but Probably talk a little bit
[7:42] Speaker 4: more about MediaTek, right?
[7:44] Speaker 2: What is MediaTek? What is it comparable to here in
[7:47] Speaker 2: the U.S.
[7:47] Speaker 7: Yeah, no, that's a very good question. This is a
[7:49] Speaker 7: very important company that nobody really pays that much attention to.
[7:53] Speaker 7: Think they're the Taiwanese equivalent of Qualcomm. They make smartphone processors,
[7:58] Speaker 7: smartphone modems. You know, the real guts of that, they
[8:01] Speaker 7: have a large presence in technology. the China market and
[8:04] Speaker 7: large presence worldwide. And by numbers, they're one of the
[8:08] Speaker 7: biggest companies. And that's really what they've done, and they're
[8:11] Speaker 7: looking to expand.
[8:12] Speaker 2: But in ASICs, they're more like Marvell or Broadcom, aren't they?
[8:14] Speaker 1: That's right. This is what they're trying to do.
[8:16] Speaker 7: They're trying to basically take that expertise that they've developed
[8:19] Speaker 7: in smartphones and expand that into the data center, which
[8:22] Speaker 7: is a more lucrative business. And what they're doing is
[8:25] Speaker 7: trying to replicate what Broadcom does, what Marvell does, which
[8:27] Speaker 7: is help, say, Google, say, Amazon. Hey, you've got this
[8:30] Speaker 7: great idea for a chip. You've got these specifications. We'll
[8:33] Speaker 7: get you all the way there.
[8:34] Speaker 4: Bloomberg's Ian King, who broke the story with me this morning,
[8:37] Speaker 4: and it's one of the most read today on the
[8:38] Speaker 4: Bloomberg platforms. Go and check out the details. NVIDIA's CEO,
[8:41] Speaker 4: Jens Wang, also talks about how the specialist platforms are
[8:45] Speaker 4: largely inference platforms. But as you heard, NVIDIA is across
[8:48] Speaker 4: every phase.
[8:49] Speaker 2: Listen to this.
[8:51] Speaker 1: An XPU is a specialized chip.
[8:54] Speaker 5: NVIDIA's GPU is a general purpose accelerator.
[8:59] Speaker 1: We accelerate the.
[9:01] Speaker 5: Entire lifecycle of AI from data processing, which data processing
[9:06] Speaker 5: to pre-training to post-training to agentic AI inferencing.
[9:11] Speaker 1: We accelerate every single AI model in the world.
[9:17] Speaker 2: Let's get more on NVIDIA.
[9:18] Speaker 4: Brad Warden, Senior Portfolio Manager at Nomura Asset Management International
[9:22] Speaker 4: joins us. And kindly, Brad, in advance, you outlined the
[9:26] Speaker 4: different phases of AI. You've heard from Jensen there about
[9:30] Speaker 4: why they did this with MediaTek to allow XPUs to
[9:33] Speaker 4: come into the ecosystem. What does that reflect and which
[9:37] Speaker 4: phase of AI you think we're in right now?
[9:40] Speaker 2: Right. Thanks, Ed. And thanks for having me this morning.
[9:42] Speaker 8: I think You know, if you think about the phases
[9:46] Speaker 8: of AI, right, we've built this foundation, right, which was,
[9:49] Speaker 8: you know, I think we started thinking about GPUs, right,
[9:52] Speaker 8: as the base of that and really saw that part
[9:55] Speaker 8: of the ecosystem develop. And then a lot of things
[9:57] Speaker 8: develop around that. And I think that's what you're talking about.
[10:00] Speaker 8: seeing right now. So you've had this foundation that's gotten
[10:03] Speaker 8: built for AI. Now we've kind of moved into this
[10:06] Speaker 8: phase of monetization. And a lot of the questions, I think,
[10:11] Speaker 8: which go back to what you were outlining earlier, just
[10:14] Speaker 8: around financing and thinking about what sort of are the
[10:18] Speaker 8: next steps of that. And I think if you really
[10:21] Speaker 8: step back and think about the relationship that was announced
[10:24] Speaker 8: that you talked about, really is this idea of basically
[10:30] Speaker 8: expanding the ecosystem and the strategic aspect of what NVIDIA
[10:35] Speaker 8: is doing. is making sure that they play across that ecosystem.
[10:39] Speaker 8: And you've seen it in a number of different investments
[10:42] Speaker 8: that they've made, but I think this one obviously extends
[10:46] Speaker 8: some of that.
[10:48] Speaker 4: Brad, I really value your perspective and reaction here because
[10:50] Speaker 4: in the Nomura Transformational Technologies ETF, the top holdings are NVIDIA, AMD, TSMC,
[10:56] Speaker 4: Seagate Land Research, but also Broadcom's in there as well, right,
[10:59] Speaker 4: on the custom ASIC side. What's the biggest motivation right
[11:03] Speaker 4: now that you're seeing from those pursuing custom silicon? The
[11:06] Speaker 4: economics on a dollar per token basis or diversity of
[11:10] Speaker 4: supply is a kind of like safety net.
[11:13] Speaker 2: That's a fantastic point.
[11:16] Speaker 8: I think if you look at each of the adoption
[11:19] Speaker 8: that each of these big hyperscalers are doing in terms
[11:22] Speaker 8: of looking at XPUs, looking at ways to really optimize
[11:26] Speaker 8: that infrastructure, which fundamentally, if you look at technology, it's
[11:31] Speaker 8: the cost per bit of output. So in this case,
[11:35] Speaker 8: it's token per watt, if you want to think about that.
[11:38] Speaker 8: And if you know your infrastructure and your operating requirements
[11:43] Speaker 8: the best out of anybody, which these hyperscalers do, their
[11:46] Speaker 8: ability to optimize that and do it in their own
[11:48] Speaker 8: silicon environment I think is something that's very, very powerful. Now,
[11:51] Speaker 8: there are a lot of reasons why you use a
[11:54] Speaker 8: general compute, a GPU as well.
[11:57] Speaker 2: And so both of those are valuable.
[11:59] Speaker 8: So how you leverage that, again, goes to optimizing that infrastructure.
[12:04] Speaker 8: And if we think about it too, recently we've been
[12:06] Speaker 8: talking about open source versus closed source models and how
[12:10] Speaker 8: those both develop. And I think the idea is that
[12:13] Speaker 8: you optimize the usage of that. And that's what technology
[12:16] Speaker 8: does is over time, we figure out ways to optimize.
[12:19] Speaker 2: And so there's a place for all of these.
[12:21] Speaker 8: And it goes back to the strategy of being able
[12:24] Speaker 8: to touch all of those points. And I think that
[12:26] Speaker 8: was the point Jensen was making in in your interview,
[12:29] Speaker 8: was really like, look, our customers are asking for all
[12:32] Speaker 8: these different architectures and capabilities. And if we can be
[12:36] Speaker 8: a partner in that and work on that, that's an
[12:38] Speaker 8: important strategic move on the part of NVIDIA to stay
[12:42] Speaker 8: at the heart of that and to make sure that
[12:44] Speaker 8: their advantage stays there. Plus, they get a view into
[12:47] Speaker 8: how the whole ecosystem is developing at every touchpoint that
[12:51] Speaker 8: they've been involved in. And that's critical from a competitive
[12:55] Speaker 8: standpoint and very smart from a strategic perspective as well.
[12:58] Speaker 2: So that ending there, interesting.
[13:00] Speaker 4: Where do you sit on the concern or lack of
[13:03] Speaker 4: concern in what is or is not circular financing in
[13:06] Speaker 4: this case?
[13:08] Speaker 2: I think it's something that's worth watching, right?
[13:10] Speaker 8: I mean, we're watching very closely the amount of leverage
[13:13] Speaker 8: that's being put on the cash flows, but we're watching
[13:16] Speaker 8: the monetization, right?
[13:17] Speaker 2: We're watching very.
[13:17] Speaker 8: Carefully the revenues that are accruing from those types of financing.
[13:22] Speaker 8: And I think the market too is clearly skeptical, right?
[13:24] Speaker 8: Because we've seen really powerful earnings power out of a
[13:28] Speaker 8: number of these companies, including the ones we've just talked about.
[13:32] Speaker 8: But really, the valuations continue to fall, right? And part
[13:36] Speaker 8: of that is that skepticism around that and the circularity.
[13:40] Speaker 2: You watch the CDS. May I just jump in and ask?
[13:43] Speaker 2: I'm sorry, because we'll run out of time.
[13:44] Speaker 4: In this MediaTek case study, the argument that both CEOs
[13:48] Speaker 4: made was, In a custom ASICs context, you can get
[13:52] Speaker 4: it into the real world quicker. You've got your custom accelerator,
[13:55] Speaker 4: you just drop it into an existing architecture. But the
[13:59] Speaker 4: counterargument to that was, well, NVIDIA's been able to find
[14:01] Speaker 4: its own demand, therefore. I'm trying to work out the
[14:05] Speaker 4: real world mechanics of how that would go.
[14:09] Speaker 2: Well, I'm not sure specifically.
[14:12] Speaker 8: I think what this does is it does definitely speed
[14:15] Speaker 8: up the capability to pull that in from a customer standpoint.
[14:19] Speaker 8: And I think that is the idea is the speed
[14:22] Speaker 8: to market and the speed with which you can connect
[14:24] Speaker 8: all these systems together is becoming more critical because we're
[14:28] Speaker 8: moving at light speed, right, in terms of how this
[14:30] Speaker 8: is happening. So as we do this, it's actually a
[14:32] Speaker 8: very smart thing.
[14:35] Speaker 2: I'm enjoying this.
[14:35] Speaker 4: Some people wake up and talk about monetary policy and
[14:38] Speaker 4: inflation expectations, and we wake up and talk about compute
[14:41] Speaker 4: and architecture. Thank you very much indeed. Numero Asset Management
[14:44] Speaker 4: International Senior Portfolio Manager, Brad Warden. Coming up, China laid
[14:49] Speaker 4: out the terms for AI talks with the US and
[14:52] Speaker 4: accused Anthropic of bad behavior.
[14:55] Speaker 2: We've got more on that story next. This is Bloomberg Tech.
[15:07] Speaker 9: It's time now for Talking Tech. I'm Yahaira Anna. And
[15:10] Speaker 9: first up, shares of Shein on the gray market are
[15:13] Speaker 9: under pressure ahead of the fast fashion giant's Hong Kong
[15:16] Speaker 9: debut Tuesday. The stock fell as much as 17% in
[15:20] Speaker 9: gray market trading after.
[15:21] Speaker 10: Its $ 1.
[15:21] Speaker 1: 7 billion IPO.
[15:22] Speaker 9: That values Shein at about $ 26 billion, well below its
[15:24] Speaker 9: nearly $ 100 billion valuation back in 2022. Plus, SK Hynix
[15:34] Speaker 9: is eyeing Japan as it races to meet booming AI demand.
[15:39] Speaker 9: The memory chip giant is studying potential manufacturing sites across
[15:43] Speaker 9: the country with its chairman, telling Bloomberg it wants locations
[15:47] Speaker 9: with good power and good water. According to a source,
[15:50] Speaker 9: a small acquisition or partnership with a Japanese company is
[15:54] Speaker 9: also on the table. And ZDOT's AI fell short of
[15:58] Speaker 9: expectations as China's AI battle heats up. First up, first
[16:03] Speaker 9: half revenue grew nearly fivefold to $ 142 million, but still
[16:08] Speaker 9: missed analyst estimates by about 30%. That's despite the company's
[16:13] Speaker 9: market value surging more than 800% since its January listing.
[16:18] Speaker 5: Ed?
[16:19] Speaker 2: Okay, thank you, Yuhaira.
[16:21] Speaker 4: China called for greater guidelines for AI models and accused
[16:24] Speaker 4: Anthropic of bad behavior ahead of AI talks with the U.S.
[16:28] Speaker 4: China said on Sunday the U.S. to prove its AI
[16:31] Speaker 4: companies are subject to the same guidelines before any, quote,
[16:34] Speaker 4: substantive talks with China could take place.
[16:38] Speaker 2: Mike Shepard joins us with the latest.
[16:40] Speaker 4: This is China trying to set the agenda, right, Shep,
[16:43] Speaker 4: ahead of what are pretty key conversations.
[16:46] Speaker 1: Very much so.
[16:47] Speaker 10: And this is a conversation, of course, that would set
[16:49] Speaker 10: the stage for a dialogue on artificial intelligence and AI
[16:53] Speaker 10: policy between President Donald Trump and Chinese leader Xi Jinping
[16:58] Speaker 10: September 24th here in Washington. And Ed, before the two
[17:01] Speaker 10: leaders sit down, though. They need to sort out exactly
[17:05] Speaker 10: what will be on the table before them. And there
[17:07] Speaker 10: needs to be this dialogue between lower-level.
[17:10] Speaker 1: Officials from the two countries.
[17:12] Speaker 10: So far, we have seen no date or agenda set
[17:15] Speaker 10: for such a conversation to take place. And we see
[17:18] Speaker 10: here in this post from a social media account that's
[17:21] Speaker 10: affiliated with the Chinese state broadcaster CCTV a version of
[17:26] Speaker 10: what it would like. Before any conversation can happen on AI,
[17:30] Speaker 10: the U.S. really does, according to China, need to meet
[17:33] Speaker 10: these terms that it is setting. American AI developers need
[17:38] Speaker 10: to be following a common set of rules when it
[17:41] Speaker 10: comes to disclosure, when it comes to audits, and when
[17:44] Speaker 10: it comes to safety. And interestingly, Ed, it also singled
[17:47] Speaker 10: out Anthropic as perhaps its main villain in the whole
[17:50] Speaker 10: opera here.
[17:51] Speaker 4: So that is interesting. I mean, they are singling out
[17:54] Speaker 4: a single U.S. company or entity. Do we know why?
[17:59] Speaker 10: Well, you know, the post itself was titled, Anthropic Has
[18:02] Speaker 10: Contracted the American Disease, and then used anthropic really as
[18:06] Speaker 10: a way of explaining all of its issues. Anthropic, though,
[18:11] Speaker 10: why them and not OpenAI, which, of course, has had
[18:14] Speaker 10: its own safety issues that we have talked about extensively
[18:17] Speaker 10: on this show with the hack on Hugging Face. Anthropic, though,
[18:21] Speaker 10: has really taken a position that is very much in
[18:23] Speaker 10: the view of officials from Beijing as anti-China. Dario Amadei
[18:27] Speaker 10: has supported the export controls that past administrations have pushed
[18:31] Speaker 10: to prevent Beijing and Chinese companies from acquiring advanced AI chips.
[18:35] Speaker 10: And he has also warned against the prospect of China
[18:38] Speaker 10: actually taking the lead in artificial intelligence development and software.
[18:43] Speaker 10: And the reason is that they see it as possibly
[18:45] Speaker 10: being used as a tool by Beijing. And of course,
[18:48] Speaker 10: we're seeing China now react against against Anthropic.
[18:53] Speaker 2: Bloomberg's Mike Sheppard with the reporting. Thank you very much.
[18:56] Speaker 4: Tomorrow, Apple gets a new CEO. But how will that
[18:59] Speaker 4: affect the iPhone makers' priorities? Bloomberg's Mark Gurman, who leads
[19:03] Speaker 4: our coverage of Apple and consumer tech, has been looking
[19:06] Speaker 4: into that and what John Turner brings to the table.
[19:09] Speaker 2: Mark is with us now.
[19:10] Speaker 4: This was the subject of this weekend's Power On. but
[19:12] Speaker 4: a really detailed piece in Bloomberg Businessweek. And let's start there. Okay,
[19:17] Speaker 4: so today is Tim Cook's last day. Tomorrow is day
[19:20] Speaker 4: one of John Ternus' CEO. What changes overnight?
[19:25] Speaker 11: I would say pretty much nothing changes overnight of any
[19:29] Speaker 11: major significance. This is going to be a gradual multi-year
[19:33] Speaker 11: changing of the guard. Tim Cook is not going anywhere.
[19:36] Speaker 11: He's still going to be around. There are questions about
[19:39] Speaker 11: how much he's going to be around. I'm led to
[19:42] Speaker 11: believe he will be very present and intimately involved with
[19:45] Speaker 11: Apple in the several years to come, serving as an
[19:48] Speaker 11: advisor to John Ternes, the new CEO, but also having
[19:52] Speaker 11: a major impact on things like supply chain, operations, procurement,
[19:58] Speaker 11: but most notably working with the Trump administration and the
[20:03] Speaker 11: government in China and other governments around the world where
[20:06] Speaker 11: he owns those hands-on relationships.
[20:09] Speaker 2: Timing is so interesting.
[20:11] Speaker 4: You know, next week, next Wednesday is the Apple full
[20:14] Speaker 4: iPhone event. How much has John Ternus had a hand
[20:19] Speaker 4: in this generation of iPhone that's likely to be announced
[20:23] Speaker 4: when we get them next week?
[20:25] Speaker 12: Well, my answer is the same for the last five generations.
[20:28] Speaker 11: John Ternus has been leading the development of the iPhone
[20:31] Speaker 11: since the release of the iPhone 12 in 2020.
[20:34] Speaker 12: He spearheaded that.
[20:36] Speaker 11: That was his final test before being named to the
[20:39] Speaker 11: CEO role earlier this year or being named to the
[20:43] Speaker 11: senior VP of hardware engineering role.
[20:45] Speaker 12: In 2021, that.
[20:47] Speaker 11: iPhone was a big success at the height of COVID
[20:49] Speaker 11: with 5G and the new design. And then they rolled
[20:52] Speaker 11: out subsequent iPhone after iPhone after iPhone under Ternus. And
[20:56] Speaker 11: the timing of the transition is not a coincidence. Typically,
[20:59] Speaker 11: they like to do these big transitions either in October, December,
[21:05] Speaker 11: or at the very beginning of the year. September 1
[21:07] Speaker 11: was picked for a very specific reason to have John
[21:10] Speaker 11: Ternus be the one to introduce the the foldable iPhone,
[21:14] Speaker 11: the smart home display, the new MacBook Pro with a touchscreen,
[21:18] Speaker 11: basically be able to present him as sort of this
[21:20] Speaker 11: new product development maestro as Apple rolls out its biggest
[21:24] Speaker 11: new product roadmap in its history.
[21:26] Speaker 4: So the thing that will change really overnight in the
[21:28] Speaker 4: internal directory at Apple is some job titles. Very quickly,
[21:32] Speaker 4: just recap what we've reported on how John Ternus is
[21:35] Speaker 4: expected to be different in his leadership, his style, I
[21:40] Speaker 4: think there's.
[21:40] Speaker 12: Going to be a lot of continuity there, right?
[21:42] Speaker 11: I mean, he has taken every opportunity to praise Cook's
[21:45] Speaker 11: leadership style. The big difference is that over the past
[21:49] Speaker 11: 15 years, decision-making at Apple has been very collaborative, where
[21:53] Speaker 11: Tim Cook has let the executive team make the decisions,
[21:57] Speaker 11: and maybe he would make the final decision, but it
[21:59] Speaker 11: was a consensus-driven leadership process. At Apple, under John Ternes,
[22:04] Speaker 11: I expect far more decisiveness and singular decision-making.
[22:08] Speaker 4: Bloomberg's Mark Gurman with two must-reads from the weekend, the
[22:10] Speaker 4: Power On newsletter and this piece in Businessweek about the transition.
[22:14] Speaker 4: Thank you very much. Generally speaking, technology stocks are lower.
[22:19] Speaker 4: There's been a flare-up in the Middle East, which is
[22:21] Speaker 4: having the sort of risk-off sentiment driven into the market.
[22:24] Speaker 2: I would say that chip stocks.
[22:26] Speaker 4: Are outperforming and modestly higher in large part because NVIDIA
[22:29] Speaker 4: has held on to the gains that it had. after
[22:32] Speaker 4: that announcement of investing in media tech. NVIDIA is making
[22:36] Speaker 4: another multi-billion dollar bet on the AI ecosystem, investing $ 3. 5
[22:39] Speaker 4: billion in what are convertible bonds in Taiwan's media tech.
[22:43] Speaker 4: I asked Jensen Wang about the proportion of the content
[22:46] Speaker 4: of the server design that NVIDIA is now responsible for,
[22:48] Speaker 4: pushing on its latest generation Vera Rubin platform and how
[22:53] Speaker 4: custom silicon would impact that.
[22:55] Speaker 2: Listen to this.
[22:56] Speaker 5: Inside a gigawatt, if we were just to extrapolate, in
[22:59] Speaker 5: a gigawatt, Hopper was about $ 18 billion of economics for
[23:04] Speaker 5: NVIDIA per gigawatt. Grace Blackwell is about 25.
[23:10] Speaker 1: And the Vera Rubin generation, it's.
[23:13] Speaker 5: About $ 40 plus billion per gigawatt of total NVIDIA compute economics.
[23:21] Speaker 5: And the reason for that is because we've been expanding
[23:24] Speaker 5: beyond the GPU to one type of switch after another
[23:29] Speaker 5: type of networking. And now the amount of chips, there's
[23:33] Speaker 5: like seven different types of NVIDIA chips that has to
[23:36] Speaker 5: go into an AI factory.
[23:39] Speaker 1: An XPU would be a replacement of one of those.
[23:41] Speaker 5: You know, my sense is that we'll continue to be
[23:45] Speaker 5: in every single cloud. We'll continue to be in every
[23:47] Speaker 5: single AI factory. Some of those racks might be augmented
[23:52] Speaker 5: with XPUs. And so in those cases, NVIDIA's networking, NVIDIA's switching,
[23:57] Speaker 5: NVIDIA's CPUs would also have an opportunity to participate. And
[24:02] Speaker 5: we make it easier for for the customers to be
[24:06] Speaker 5: able to extend out their AI factory using exactly the
[24:09] Speaker 5: same system architecture and networking architecture. In the case of MediaTek,
[24:14] Speaker 5: it allows MediaTek to offer an end-to-end solution that is complete.
[24:19] Speaker 5: That XPU otherwise would have to go and solve MV-Link
[24:23] Speaker 5: switching problems, a whole bunch of NICS problems, a whole
[24:26] Speaker 5: bunch of Ethernet switching.
[24:28] Speaker 1: Scale-outs, scale-up, scale across, scale in.
[24:33] Speaker 5: There's so many different types of networking technologies that have
[24:36] Speaker 5: to come to bear, and we offer the entire thing.
[24:40] Speaker 5: So between us and MediaTek, if a customer wants to
[24:42] Speaker 5: build XPUs, they now have an entire solution as well.
[24:46] Speaker 5: And so our opportunity will continue to grow in these
[24:50] Speaker 5: AI factories.
[24:53] Speaker 4: That was NVIDIA CEO Jensen Wong earlier this morning. Anthropix
[24:56] Speaker 4: IPO plans are casting a bit of a shadow over
[24:59] Speaker 4: other companies' U.S.
[25:00] Speaker 2: Listing plans.
[25:01] Speaker 4: The Claude developer is preparing to file publicly for an
[25:03] Speaker 4: offering that is expected to be as large or larger
[25:07] Speaker 4: than SpaceX's record $ 86. 2 billion debut. Bloomberg Equity's reporter, Bailey Lipschultz,
[25:13] Speaker 4: is here. ECM, IPO czar. It's getting kind of real.
[25:18] Speaker 4: And I find this so interesting that in our reporting,
[25:21] Speaker 4: It's not just us waiting for Anthropic to flip the
[25:25] Speaker 4: S1 public.
[25:26] Speaker 2: It's everyone else.
[25:27] Speaker 4: Waiting for them to do it as well to decide
[25:29] Speaker 4: what it is they want to do with their own
[25:30] Speaker 4: public market futures.
[25:32] Speaker 3: Well, Ed, the thing you have to keep in mind
[25:34] Speaker 3: too is we have a very much live Fed meeting
[25:37] Speaker 3: in the middle of September. And then we have an
[25:39] Speaker 3: election that gets remarkably close the longer Anthropic drags its heels.
[25:43] Speaker 3: So when I'm talking to bankers, private equity sponsors, and
[25:46] Speaker 3: pretty much everyone in the investing world, they're kind of wondering, okay,
[25:49] Speaker 3: when will this flip public? Once we have an idea
[25:52] Speaker 3: on when they're going public, we either have to, A,
[25:54] Speaker 3: hurry up and get out the door or prepare to
[25:56] Speaker 3: get out the door. We're looking at potentially seeing filings
[25:58] Speaker 3: from the likes of an N-scale, an Aura, and an
[26:01] Speaker 3: SB Energy. Or you've got to sit on your hands,
[26:03] Speaker 3: let Anthropic do its thing, and hope the market is
[26:06] Speaker 3: still there in the aftermath. And that was a debate
[26:08] Speaker 3: we saw with SpaceX.
[26:10] Speaker 4: So there's been even just in the last five days,
[26:12] Speaker 4: like lots of anthropic headlines and it's dominated the news
[26:16] Speaker 4: cycle even before it's happened, the S1 flipping public. And
[26:19] Speaker 4: in your report, you detail that like a lot of
[26:22] Speaker 4: those other IPOs or potential IPOs that you're mentioning are
[26:25] Speaker 4: finding it hard to get the attention of some investors.
[26:29] Speaker 3: Well, and the thing you have to keep in mind, too,
[26:31] Speaker 3: is the investors that a lot of these companies want
[26:34] Speaker 3: one on ones with sovereign wealth funds and long only investors,
[26:37] Speaker 3: they're not big teams. So if you were trying to
[26:40] Speaker 3: court a Wellington or a Norges, you kind of have
[26:44] Speaker 3: to deal with, OK, well, what happens if Dario comes
[26:46] Speaker 3: calling because they're going to cancel any meeting that they have?
[26:49] Speaker 3: to meet with the team from Anthropic, even if they
[26:51] Speaker 3: are invested in them privately. So when we talk to
[26:53] Speaker 3: some of these sponsors, when we talk to some of
[26:55] Speaker 3: these bankers and these VCs, the question comes back to,
[26:58] Speaker 3: we don't want to be on the road when Anthropic
[27:00] Speaker 3: is on the road. We don't want Anthropic to have
[27:02] Speaker 3: an investor day that all of a sudden now Wednesday,
[27:06] Speaker 3: pick a date in September is now off the table
[27:08] Speaker 3: because everyone is going to be tuned into that call.
[27:11] Speaker 3: And then immediately after that, having internal discussions and talking
[27:13] Speaker 3: to everyone they know on the street. So it is
[27:16] Speaker 3: potentially going to impact what this race to kind of
[27:20] Speaker 3: the end of October looks like from an IPO perspective.
[27:23] Speaker 3: So Bailey, let's quickly end by recapping the reporting. Where
[27:26] Speaker 3: do we stand with what we think will happen with
[27:28] Speaker 3: Anthropic's IPO? We're expecting Anthropic to finalize its revolving credit facility.
[27:32] Speaker 3: That's the first big step, the next big step. After that,
[27:35] Speaker 3: banks come in under the tent. They're able to kind
[27:37] Speaker 3: of get their hands on the numbers. They'll have some
[27:40] Speaker 3: form of an analyst day. Then the public flip. So
[27:43] Speaker 3: public flip, maybe the end of this week. On the
[27:45] Speaker 3: other side of Labor Day, that kickstarts the two-week waiting period.
[27:48] Speaker 3: Then you have the launch deal pricing as early as
[27:51] Speaker 3: the end of September. But if you look at the
[27:53] Speaker 3: calendar right now, things really have to get in motion
[27:55] Speaker 3: for it to not slip into October.
[27:59] Speaker 2: Thank you very much. We have some interesting data here.
[28:02] Speaker 4: Madrona has released its sixth annual IA40 list of the
[28:06] Speaker 4: top private companies building and enabling AI. Madrona is an
[28:10] Speaker 4: investor in the likes of OpenAI and Anthropic, but it
[28:13] Speaker 4: sources nomination and votes for this list from 72 outside
[28:17] Speaker 4: investors across 54 firms and evaluates the list against PitchBook data.
[28:22] Speaker 4: Matt McElwain, Managing Director of Madrona, is with us. I
[28:24] Speaker 4: find this different. It's a different approach to trying to
[28:27] Speaker 4: find out real world what's happening right now. Just explain
[28:32] Speaker 4: the methodology and the list itself.
[28:35] Speaker 13: Yeah, so this is the sixth year of the list,
[28:37] Speaker 13: and it's the fifth year of the summit that we
[28:40] Speaker 13: host every year. And what we have done is we
[28:42] Speaker 13: have surveyed the entire market of venture capitalists and technology
[28:46] Speaker 13: leaders at companies like Microsoft and Amazon and others to
[28:50] Speaker 13: give us their best nominations. And then we have a
[28:52] Speaker 13: voting mechanism and then we combine it with pitch book
[28:54] Speaker 13: data and we come up with this list. Here's what's
[28:57] Speaker 13: amazing about the list. The first year we launched it
[29:00] Speaker 13: in 2021, pre-ChatGPT. The entire group of winners raised $ 15 billion.
[29:07] Speaker 13: This year, the list will have raised over $ 300 billion.
[29:11] Speaker 13: And in aggregate, over the last five years, those companies
[29:16] Speaker 13: have raised over $ 500 billion. So as you're talking about
[29:19] Speaker 13: companies like Anthropic and its IPO, and Anthropic is on
[29:23] Speaker 13: the list, as is OpenAI, those are companies that have
[29:26] Speaker 13: already raised approaching $ 200 billion in the private market before
[29:30] Speaker 13: they ever went public.
[29:32] Speaker 1: So these are really the leading indicators of companies.
[29:35] Speaker 13: Much more in the applied space and in the agentic
[29:38] Speaker 13: space relative to the intangible. space that are going to
[29:42] Speaker 13: be the next generation of leaders.
[29:44] Speaker 4: There are several conclusions you can draw, and it's several
[29:47] Speaker 4: different ways to slice the data, right? So I think
[29:51] Speaker 4: the 45 winners, $ 410 billion between them raised since founding.
[29:55] Speaker 1: Correct.
[29:56] Speaker 4: But of that $ 410 billion, just three names account for
[29:59] Speaker 4: most of it.
[29:59] Speaker 13: The three names, no surprise, OpenAI, Anthropic, and Databricks.
[30:03] Speaker 1: And Databricks.
[30:04] Speaker 13: That's $ 375 billion roughly of all of that capital. But
[30:07] Speaker 13: what's interesting is that still leaves $ 800 million per company
[30:13] Speaker 13: for all those other companies that have ever been on
[30:15] Speaker 13: the list.
[30:16] Speaker 1: So a tremendous amount of money. And remember that we.
[30:19] Speaker 13: Restrict the early and mid-stage to companies that have raised
[30:23] Speaker 13: less than $ 200 million. And so over time, many companies
[30:26] Speaker 13: have migrated from being in the early and mid-stage of
[30:29] Speaker 13: size of capital that they've raised up into the larger amounts.
[30:33] Speaker 13: What's also interesting is the turnover on this list. Roughly 50%
[30:38] Speaker 13: a year, some years even higher, of companies that were
[30:40] Speaker 13: on the list the year before fall off the list
[30:43] Speaker 13: in the following year, which goes to this question of
[30:46] Speaker 13: durability versus disruption. We all see tons of disruption going on,
[30:50] Speaker 13: but what companies can be durable and how can they
[30:53] Speaker 13: be durable?
[30:55] Speaker 4: When I look at the list, I want to dwell
[30:57] Speaker 4: just a little bit on those three, but principally OpenAI
[31:01] Speaker 4: and Anthropic. It also is a reminder of where the
[31:03] Speaker 4: capital's coming from. And in the case of the Frontier Labs,
[31:07] Speaker 4: it's what people call strategics. But the hyperscalers and NVIDIA themselves,
[31:10] Speaker 4: obviously you heard some of our interview with Jensen Wong
[31:13] Speaker 4: and NVIDIA making a pretty sizable investment in media tech.
[31:17] Speaker 4: What do we infer from that? That at this scale,
[31:21] Speaker 4: that's where the money's coming from, as opposed to venture capital.
[31:24] Speaker 1: Well, it is more from strategics now.
[31:27] Speaker 13: And what I say to both the big tech companies
[31:29] Speaker 13: and to the people funding the model companies, thank you
[31:33] Speaker 13: very much. We really appreciate, you know, the five largest companies,
[31:37] Speaker 13: you know, in technology are You know, in the world,
[31:40] Speaker 13: the Amazons, Microsofts, Googles will combine, invest an estimated $ 750
[31:46] Speaker 13: billion in.
[31:46] Speaker 1: CapEx this year.
[31:48] Speaker 13: On top of that, you've got hundreds of billions of
[31:50] Speaker 13: dollars going into open AI and Anthropic. But what is
[31:53] Speaker 13: the ROI of AI? That is the core question that's
[31:57] Speaker 13: increasingly being asked. by enterprises, by the businesses that leverage
[32:02] Speaker 13: all these tokens. You've heard about token maxing. You've heard
[32:04] Speaker 13: of all these different dynamics of, am I getting an
[32:07] Speaker 13: ROI as an end-user customer as a.
[32:10] Speaker 2: Is there an answer to that question yet?
[32:12] Speaker 13: Oh, I think that you look through this list and
[32:14] Speaker 13: you're going to see companies that are not only generating
[32:18] Speaker 13: good gross margins, a lot of the companies like the
[32:21] Speaker 13: Servals and Gradials and Vercels that are on this list,
[32:24] Speaker 13: but they are delivering a ton of value to their
[32:27] Speaker 13: end-user customers.
[32:28] Speaker 1: Whether those customers are in.
[32:29] Speaker 13: A specific functional area, like marketing or IT or sales,
[32:34] Speaker 13: or they're more cross-cutting in serving builders, as in the
[32:37] Speaker 13: case of Averse Cell and others.
[32:39] Speaker 4: The other domain that comes to the forefront here is coding.
[32:44] Speaker 4: You talked a little bit about some of the players
[32:45] Speaker 4: in that field, but it's interesting to see so many
[32:49] Speaker 4: of those involved in coding make the list. That is
[32:52] Speaker 4: on a daily basis on this program where the news
[32:55] Speaker 4: flow is at the moment.
[32:56] Speaker 1: Well, I think coding is very important.
[32:58] Speaker 13: And Cursor, who was on the list for many years,
[33:00] Speaker 13: they've now been acquired, of course, by SpaceX.
[33:04] Speaker 1: Led the way.
[33:05] Speaker 13: But you really should think of it more broadly as builders.
[33:08] Speaker 13: How are we enabling builders in so many different areas?
[33:11] Speaker 13: A company like Lovable that is in the vibe coding
[33:13] Speaker 13: area of building, or again, a company like Gradial that's
[33:17] Speaker 13: helping marketers and the enterprises, the Global 500, be able
[33:22] Speaker 13: to build amazing solutions in applied marketing areas. So I
[33:26] Speaker 13: think it's about builders. I think it's about collaboration, not
[33:30] Speaker 13: just between humans, but humans and agents. And then finally,
[33:33] Speaker 13: I think it's about the continuous learning loop. You're not
[33:36] Speaker 13: hearing as much about this yet. We are hearing about
[33:39] Speaker 13: it as early stage venture capitalists, that continuous learning from
[33:43] Speaker 13: taking the engagement I have with builders, running that data
[33:47] Speaker 13: through models and systems and And coming out with continuously
[33:50] Speaker 13: improving systems like Cursor did, which is why Cursor has
[33:53] Speaker 13: been so successful.
[33:54] Speaker 4: Matt, you heard Bailey's reporting there very quickly. What does
[33:57] Speaker 4: an anthropic IPO do to this list and this ecosystem?
[34:00] Speaker 13: Well, I think it actually clears out the way for
[34:04] Speaker 13: future IPOs.
[34:05] Speaker 1: You know, I think there's a number of companies on
[34:08] Speaker 1: the list.
[34:09] Speaker 13: That are going to wait for that anthropic IPO, not
[34:12] Speaker 13: the least of which, of course, OpenAI. But you look
[34:15] Speaker 13: down the list and you see some really great companies
[34:17] Speaker 13: like a ClickHouse, like a Databricks and others. And I
[34:20] Speaker 13: think if the anthropic IPO is successful, we can all
[34:24] Speaker 13: debate what the definition of a successful IPO is, but
[34:27] Speaker 13: let's say it's successful and it's viewed as successful in
[34:29] Speaker 13: the capital markets. I think you'll see that be an
[34:31] Speaker 13: opening of the gates in 2027. Matt McElwain, Managing Director
[34:35] Speaker 13: for Madrona. I've really enjoyed having you here. Thank you
[34:37] Speaker 13: very much indeed. Coming up, the U.S.
[34:39] Speaker 4: Army tapped five nuclear reactor suppliers in an effort to
[34:43] Speaker 4: boost America's nuclear industry. Radiant President and Chief Operating Officer
[34:46] Speaker 4: Tori Shavander joins us next.
[34:49] Speaker 2: This is Bloomberg Tech.
[35:01] Speaker 4: Okay, some breaking news on the Bloomberg terminal. Tim Cook
[35:03] Speaker 4: says he has enormous comfort with John Ternus taking the
[35:07] Speaker 4: helm as CEO and said that he will miss this
[35:10] Speaker 4: work as he has his final day as CEO of Apple.
[35:14] Speaker 4: This was in a email that he sent to all
[35:17] Speaker 4: Apple employees this Monday, again, his final day as CEO.
[35:21] Speaker 4: He writes, as you know, I am not leaving Apple,
[35:23] Speaker 4: but I'm stepping away from a role that I have
[35:26] Speaker 4: loved deeply. He continues in the email to say, I
[35:29] Speaker 4: will miss this work in ways I can only begin
[35:31] Speaker 4: to imagine, even as I remain completely at peace with
[35:35] Speaker 4: my decision. Cook said that a few people understand what
[35:40] Speaker 4: it takes to build products that change the world the
[35:42] Speaker 4: way that John Ternus does. We were earlier in the
[35:44] Speaker 4: show talking to Bloomberg's Mark Gurman, who detailed the transition
[35:48] Speaker 4: in Bloomberg Businessweek over the weekend. Tomorrow, John Ternus becomes
[35:52] Speaker 4: CEO of Apple and Tim Cook moves to the executive
[35:55] Speaker 4: chair role. If we hear any more, we'll bring it
[35:57] Speaker 4: to you.
[35:58] Speaker 2: Okay, the U.S.
[35:59] Speaker 4: Military recently tapped five suppliers, investing $ 2. 2 billion to boost
[36:05] Speaker 4: the nation's nuclear industry. Radiant was among the five firms tapped,
[36:11] Speaker 4: earning a $ 750 million contract to provide 15 nuclear micro-reactors.
[36:18] Speaker 4: Radiant President and Chief Operating Officer Tori Shibandon joins us now.
[36:24] Speaker 2: Really interesting.
[36:25] Speaker 4: Last week, this was a big story, a big story
[36:27] Speaker 4: about America taking action for the technology in the field,
[36:30] Speaker 4: but also to see the work of the companies doing it.
[36:33] Speaker 4: Could we just start by explaining, Tori, what is Radium?
[36:37] Speaker 4: What is its point of difference in the field of
[36:40] Speaker 4: those working on the technology?
[36:42] Speaker 2: Yeah, absolutely.
[36:43] Speaker 14: Thank you very much for having me. So Radiant builds
[36:46] Speaker 14: one megawatt microreactors that are transportable by land, air, and sea.
[36:51] Speaker 14: One megawatt is about enough power for a thousand homes.
[36:54] Speaker 14: Our big differentiator is that we are portable. And last
[36:57] Speaker 14: week we were awarded $ 750 million for 15 of our
[37:01] Speaker 14: portable reactors for the Army Janus program. That's about one
[37:05] Speaker 14: third of the overall Janus program budget, but also about 75%
[37:09] Speaker 14: of the overall Janus program volume. So 50 of 20
[37:13] Speaker 14: reactors going portable, really showing that a vote of confidence
[37:17] Speaker 14: in the maturity of our tech, as well as our
[37:19] Speaker 14: ability to scale and manufacture.
[37:21] Speaker 2: I'd like to thank Dr.
[37:22] Speaker 14: Jeff Waxman and DIU Director Owen West for their bold
[37:26] Speaker 14: leadership and understanding that portable nuclear is the big solution
[37:30] Speaker 14: that the Army needs.
[37:32] Speaker 2: Tori, what did you have to demonstrate in this process
[37:35] Speaker 2: to win the award?
[37:36] Speaker 14: Yeah, so it's a pretty grueling process through multiple down selects.
[37:41] Speaker 14: We've competed against, you know, over 40 companies in this.
[37:44] Speaker 2: And the U.S.
[37:46] Speaker 14: Military is an incredibly savvy customer. They have many experts
[37:50] Speaker 14: where we've been through hours of presentations as well as
[37:53] Speaker 14: grilling on our understanding of what our product is capable
[37:56] Speaker 14: to do, as well as them coming on site to
[37:58] Speaker 14: see the technology we've developed. We shipped recently our test
[38:03] Speaker 14: reactor up to the Idaho National Lab dome where it
[38:05] Speaker 14: will begin our overall testing over the next few months
[38:10] Speaker 14: because the next reactor that we will deliver to a
[38:12] Speaker 14: military base in 2028, a customer really needs us to
[38:16] Speaker 14: understand our product and the military's selection of us in
[38:20] Speaker 14: this award is a vote of confidence.
[38:23] Speaker 4: Tori, when your industry comes on the program, almost universally
[38:27] Speaker 4: they talk about China and how China is either ahead
[38:29] Speaker 4: on the regulation side or on the technology side.
[38:32] Speaker 2: What is Radiant's house view on that, please?
[38:35] Speaker 14: Yeah, so when I look at what Radiant's solving, we're
[38:38] Speaker 14: solving the entire problem of the American energy grid. And
[38:41] Speaker 14: when you compare it to China, we're solving different problems
[38:45] Speaker 14: because our baseline solution is different. The American energy grid
[38:49] Speaker 14: was built in the 60s. It hasn't really changed. We
[38:52] Speaker 14: have generation, transmission, and utilization. And what Radiant is solving
[38:57] Speaker 14: is being able to solve generation and transmission on the
[39:00] Speaker 14: customer site together. You can either just build large CapEx
[39:04] Speaker 14: projects fast, that's what China is approaching, or you can
[39:07] Speaker 14: be innovative. This is what America and the American military
[39:10] Speaker 14: are best at and come up with a new solution
[39:13] Speaker 14: to make a more resilient grid to build the future
[39:15] Speaker 14: America needs.
[39:16] Speaker 2: Tori, very quickly, what do you need to do next
[39:19] Speaker 2: to scale?
[39:19] Speaker 4: Will you raise money? Will you hire competitively?
[39:23] Speaker 14: No, we will do all of the above, but most importantly,
[39:25] Speaker 14: we'll put our heads down and build for America. What
[39:29] Speaker 14: folks need to understand is that our grid needs a
[39:32] Speaker 14: new solution and portable nuclear is that new solution.
[39:36] Speaker 2: Radiant President and Chief Operating Officer, Tori Shivanandan. Great to
[39:40] Speaker 2: have you on the show. Thank you very much.
[39:42] Speaker 4: Let's get back to our top story, NVIDIA investing $ 3. 5
[39:45] Speaker 4: billion in Taiwan's MediaTek. Bloomberg Intelligence saying this gives NVIDIA, quote,
[40:01] Speaker 4: With more, it's Kunjan Sabani of Bloomberg Intelligence who wrote
[40:03] Speaker 4: that note. So interesting. Custom silicon in conjunction with NVIDIA's
[40:09] Speaker 4: GPU-based architecture. Just take us deeper into the React, Kunjan.
[40:13] Speaker 15: Yeah, so most of the times when we discuss what's
[40:15] Speaker 15: the biggest challenge to NVIDIA GPUs, it's these custom silicon, right?
[40:19] Speaker 15: And this is a great hedge that NVIDIA is playing.
[40:21] Speaker 15: So even in a data center that is built up,
[40:24] Speaker 15: not using NVIDIA's GPUs, but using custom silicon, like whether
[40:28] Speaker 15: it's a TPU, whether it's an Amazon system silicon, These
[40:31] Speaker 15: kind of partnerships will allow NVIDIA to stay within the stack,
[40:36] Speaker 15: both from a hardware and a software perspective, and hence
[40:39] Speaker 15: not lose out on the entire CapEx dollar spent.
[40:42] Speaker 16: We have already seen a glimpse of this play out.
[40:45] Speaker 15: NVIDIA had a similar arrangement with Marvell, which is the
[40:47] Speaker 15: key ASIC provider for AWS.
[40:49] Speaker 1: Yes.
[40:50] Speaker 15: And we are seeing this NVLink Fusion deployment happening in AWS.
[40:54] Speaker 16: And guess what?
[40:54] Speaker 15: Recently, AWS has been the biggest NVIDIA GPU customer. So
[40:59] Speaker 15: read what you will in there, but this is definitely
[41:01] Speaker 15: a great win-win situation.
[41:03] Speaker 4: So the stock's higher for NVIDIA. I don't know that
[41:06] Speaker 4: I'm going to draw a direct causal link. The most
[41:08] Speaker 4: interesting point in the interview was probably what Rick Zai said,
[41:11] Speaker 4: which is, you know, XPUs are difficult, but this actually
[41:14] Speaker 4: accelerates speed to deployment by doing it this way with
[41:17] Speaker 4: NVIDIA NVLink.
[41:18] Speaker 2: Do you buy that argument?
[41:22] Speaker 16: Somewhat true.
[41:23] Speaker 15: Look, the challenge that XPUs have, especially when deploying it
[41:26] Speaker 15: to external parties, it's a very closed ecosystem, both from
[41:30] Speaker 15: a software perspective and a network perspective. So this does
[41:33] Speaker 15: help them with time to market. But I would not
[41:36] Speaker 15: stretch it to a degree and say this really helps
[41:39] Speaker 15: them become sort of a chip seller to the whole industry.
[41:43] Speaker 4: Let's do the kind of finance bit. NVIDIA buys $ 3. 5
[41:47] Speaker 4: billion of convertible bonds of MediaTek. It's their biggest non-US
[41:52] Speaker 4: direct investment. What does BI think about that mechanism?
[41:57] Speaker 15: I think, you know, like I mentioned in my prior interview,
[42:00] Speaker 15: this is sort of the bank of Jensen or the
[42:02] Speaker 15: bank of Nvidia really flexing its muscle.
[42:04] Speaker 16: We see these as great use of cash.
[42:07] Speaker 15: I mean, look, the amount of cash they are generating,
[42:10] Speaker 15: what can they go ahead and do, right? Last time
[42:12] Speaker 15: when they announced an increase in dividend, the street didn't
[42:15] Speaker 15: really like it.
[42:16] Speaker 16: It took it as a signal that the growth is slowing.
[42:18] Speaker 15: Even if they do hundreds of billions of buyback, that's
[42:21] Speaker 15: basically single digit percentage buyback for the stock. They are
[42:24] Speaker 15: already investing heavily in R & D and OPEX. So
[42:28] Speaker 15: we think this use of cash is definitely going to
[42:30] Speaker 15: pay a lot more ROI versus the other use of cash.
[42:34] Speaker 4: Kunjan Svani of Bloomberg Intelligence with a react that if
[42:37] Speaker 4: you are a terminal subscriber or client, go and read
[42:40] Speaker 4: the research, super important, and watch the interview back.
[42:43] Speaker 2: It was very interesting. That does it for this edition
[42:45] Speaker 2: of Bloomberg Tech.
[42:46] Speaker 4: This is what markets look like, some anxiety in the
[42:49] Speaker 4: Middle East with Iran, probably the biggest factor, some outperformance
[42:53] Speaker 4: in chips because of NVIDIA. Recap that interview, or at
[42:56] Speaker 4: least some of it, on the podcast. You know where
[42:57] Speaker 4: to find it. It was a 20-minute conversation, so you
[43:00] Speaker 4: can also check out the full version wherever you get
[43:02] Speaker 4: your interviews.
[43:03] Speaker 2: This is Bloomberg.
[43:05] Speaker 13: Thank you.

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.