I'm wondering if this change in leadership at NVIDIA and the resignation of Huang might reshape any of those dynamics or lead to questions about, you know, whether China is going to pull ahead.

Well, the U.S. and China have been competitors at the government level, but NVIDIA has always straddled the line.

Most of its customers have been in the United States, but everyone has known that Chinese AI companies like DeepSeek or Alibaba have used smuggled NVIDIA chips that arrived into China to train their own customers.

AI systems.

And for NVIDIA, I think one question that will be asked is how much exactly did they know about that smuggling and what was NVIDIA doing to prevent this? These are the types of questions that I think Congress is going to ask of NVIDIA over the coming

Can you spell out a little bit more what the national security concerns are about having companies like that have access to these NVIDIA chips?

Well, I think there are two categories of concern.

First is that there's no doubt that AI is going to be one of the most important technologies over the next couple of decades and that AI companies will be critical nodes in the world economy.

I'm wondering if this change in leadership at NVIDIA and the resignation of Huang might reshape any of those dynamics or lead to questions about, you know, whether China is going to pull ahead.

Well, the U.S. and China have been competitors at the government level, but NVIDIA has always straddled the line.

Most of its customers have been in the United States, but everyone has known that Chinese AI companies like DeepSeek or Alibaba have used smuggled NVIDIA chips that arrived into China to train their own AI systems.

And for NVIDIA, I think one question that will be asked is how much exactly did they know about that smuggling and what was NVIDIA doing to prevent this? These are the types of questions that I I think Congress is going to ask of Nvidia over the coming

Can you spell out a little bit more what the national security concerns are about having companies like that have access to these Nvidia chips?

I think there are two categories of concern.

First is that there's no doubt that AI is going to be one of the most important technologies over the next couple of decades and that AI companies will be critical nodes in the world economy.

And the U.S. wants U.S. companies out in front.

And like, I don't, you know, I just don't know where the bottom is.

Yeah, I think that all sounds right.

And I think if you look at the response in Washington, it's going to be to put tougher controls on Nvidia and tougher controls on the Trump administration to make them enforce the rules that are actually on the book.

And so it's an interesting question for China.

Are the second order effects negative rather than positive? can no longer smuggle as many chips in through Southeast Asia or get access to as many data centers that are in turn training and deploying Chinese models.

It'll be an important question to see what, if anything, China did learn as a result of this.

Right now, too soon to tell.

Does that track with your understanding of how the the global political business landscape has evolved?

But tell me, why is it that chips are so hard to make and why aren't they made in the United States more?

Well, these tiny transistors that you reference are today microscopic in size, measured in just a handful of nanometers.

That's billionths of a meter in scale, yet they're produced by the trillions and trillions.

And so there's nothing that comes close to the complexity of manufacturing of these chips.

And today there's just one company in the world that actually does the manufacturing.

NVIDIA itself, like many chip firms, only actually designs the semiconductors.

And that's a very hard thing to do on its own, but it works closely hand in hand with the Taiwan Semiconductor Manufacturing Corporation, which manufactures almost all of the world's AI semiconductors.

Why doesn't China have the ability to build more of these chips?

Getting access to these Nvidia chips, what would that mean?

right now the united states and its close allies korea and taiwan produce the vast majority of the world's ai chips the semiconductors that make ai systems possible something like 97 or 98 percent of the world's ai chips are produced outside of china and because of that china doesn't have access to enough chips and it either needs to smuggle in physically smuggle in nvidia chips into china which happens at substantial scale, or it needs to access them remotely via data centers in other countries.

And this is where Malaysia and Thailand have become locations for Chinese firms to strike deals with local data center operators and get access to NVIDIA's chips.

And this is critically important because all of the world's AI models, even Chinese models, are almost exclusively trained on NVIDIA semiconductors.

But what should they be doing?

Well, I think clearly the board of directors needs to bring in some outside expertise to help navigate this.

Jensen is, of course, a unique figure in the AI world.

But I think it's fair to say that NVIDIA's leadership ranks have just been insufficiently focused on the ways in which geopolitics is playing a central role in the chip industry.

How should we think about critical minerals with regards to semiconductor chips and this industry?

What's a key question? China is the world's leading producer of rare earths, of both gallium, germanium, indium, tungsten, all of which are needed in the production of semiconductors, either the chips themselves or in the supply chain that's used to produce chips.

When you dig into the materials that undergird the chip industry, there's a lot of unique capabilities, advanced materials that only a small number of companies can produce.

So if you look at the materials, for example, in advanced packaging, bringing different chips together in different places, Japanese companies have really unique capabilities and ultra precise production of these materials.

Actually, you think that advanced packaging is one of the key checkpoints right now? It certainly is.

There's tons of investment going into the space just as we speak.

What are the changes you're seeing there at the moment? So first, we're stacking chips on top of each other in new ways.

What about AI applications now? Where do you think we're going to see the major increases that's going to impact compute demand?

D- just help us, as you have with these other questions, think through this debate about, about the competition with China and the role that this plays.

So I think there are a couple different aspects to the, the competition.

There's the, first, the race to build the AI platforms that everyone else is gonna rely on, so companies like OpenAI, Anthropic, and others.

And it's very much in our interest to have the world operating as much as possible on US AI infrastructure because then we get to set the rules and, and China doesn't.

And you can imagine a world in which China gets to set the rules of all the AIs that we rely on, be a much less attractive place to live.

So that, that's one.

Second, there's a race to deploy AI as rapidly as possible because ultimately we, we don't only want great AI companies, we want every company to be deploying AI to become more efficient, build better products, devise better medicines, et cetera.

Does that give Taiwan an asymmetric advantage over, over Beijing, and both militarily and economically going forward, and how will that play out?

D-just help us, as you have with these other questions, think through this debate about, about the competition with China and the role that this plays.

So I think there are a couple different aspects to the, the competition.

There's the, first, the race to build the AI platforms that everyone else is gonna rely on, so companies like OpenAI, Anthropic, and others.

And it's very much in our interest to have the world operating as much as possible on US AI infrastructure because then we get to set the rules and, and China doesn't.

And you can imagine a world in which China gets to set the rules of all the AIs that we rely on, be a much less attractive place to live.

So that, that's one.

Second, there's a race to deploy AI as rapidly as possible because ultimately we, we don't only want great AI companies, we want every company to be deploying AI to become more efficient, build better products, devise better medicines, et cetera.

Does that give Taiwan an asymmetric advantage over, over Beijing, and both militarily and economically going forward, and how will that play out?

You were right on the precipice of being the first of the punk bands, but I think it all started with opening for the Pistols, is that right?

Well, w- we only ever did two shows for the Pistols.

One was our very first gig.

There was maybe 50 people there, if that.

And, um, they were playing the 100 Club.

The 100 Club was sympathetic with having those kind of bands on, 'cause we didn't have a name then.

We weren't punk bands, we were just bands that were different.

Did you realize that you were groundbreaking at the time? Or you, you were just going against what had come before you, specifically trying to do something different?

Yeah

... unfold it and put it on the, on the floor, and I'm circling job ads and what am I gonna do, and I've...

You know, one of them must have been a strata management job that I circled, and I went to it.

I hadn't got back to the car after the interview where I got the phone call saying, "You got the job." I f- had no idea what a strata manager was and, um, that was, you know, 20-something years ago now, and I just...

I, I don't know, but it took me from being fairly rudderless to sort of going, "Oh, yeah, property strata management.

This is hard yakka," but it was sort of interesting.

I sort of thought it could be a springboard to other things.

[laughs]

But can you sort of give us some sense of h- of firstly how to think about the value chain of the creation of a semiconductor from design, the tools, lithography, and then can you s- I guess tell us how, I think probably the easiest way to talk about this is TSMC in Taiwan, give us some sense of h- of how and why it ended up that very few US chip companies had their own fabrication capabilities.

To start with the value chain, i- if you want to make a semiconductor today, you, you need to begin by acquiring design software, software that's, uh, that's, uh, specialized in chip design.

If you wanna lay out 10 billion transistors on a piece of silicon the size of your fingernail, you need very automated processes to do so.

There's basically just three companies that have, uh, an almost oligopolistic position in chip design software, all which are based in the US.

Then you actually need to design the chip to do what you want it to do, and there's lots of different chips that do different things.

Some chips manage Wi-Fi connections, other chips train AI systems, other chips remember data.

E- each of these types of designs involves a lot of really unique knowhow, um, such that there's only, in most cases, a couple of companies that can design chips at the cutting edge in each of these different spheres, and a lot of unique process knowledge that's, uh, retained inside of these companies as well as a lot of unique intellectual property.

Can you help us untangle, you know, the subsequent CHIPS Act in the US, the, the, uh, trade wars that, uh, you know, the, the, the entity list, the banning of, of, of key technology going to China, particularly, you know, these latest GPU chips that are crucial for AI? Can you just help us understand that whole landscape taking us up to where we are today?

If you don't mind, just tell the audience a little bit about your background, how you got into finance and tax solutions, and just take us to how you got to where you are today and give us some context for the conversation.

Yeah.

So our story, you know, really, I guess it originates with my father.

He's my original partner, but he sort of semi-retired these days, kind of just works with a handful of clients.

But I was I had the privilege of getting out during the recession of 08 09.

So there weren't a lot of job opportunities.

no matter what your degree was.

They don't know how to handle the taxes.

But for those who are watching this for the first time, tell us about who the

HubCorp is and what they do.

Sure.

So the HubCorp, we're a manufacturer, and our purpose is to challenge the standards.

We make aftermarket parts, unlike parts that you've seen before.

Parts that are more durable, that have more functionality, that drive performance, that are easier to service.

And of course,

So my question for you, Chris, is what are you excited about when you think about someone like Paul who's built this great business and is now going to be distributing your product? What are you excited about

Um, but, but w- what do you see as sort of, like, the key inflection points that people should be watching for as the semiconductor industry continues to boom or, you know, tops out at some point?

Yeah.

I, I think right now it is, it is gated by TSMC and manufacturing capacity more broadly.

But if you put yourself in the shoes of those companies, they're looking at AI demand, not sure how strong it's gonna be.

Obviously strong, but is it strong enough to double a tripling of semiconductor spend as a share of GDP? That's a, a big jump.

And of course, they're on the hooks for the CapEx.

They gotta build the factories, they gotta pay for them.

Yeah.

How has Ukraine been able to ramp up its drone production on such a large scale?

Well, in the sort of spirit of mass mobilization here, companies that made household appliances or companies that were involved in the steel industry have now refurbished their production lines to produce parts for weapons systems and drones.

There are Saturday night house parties now where people will get together and make dinner and wine and then put together FPV or first person view drones.

And so we're seeing it done on small hobby scales, but also on this, uh, greater industrial scale.

Thanks not only to its resources and capacity here, but to Western partners, particularly Europe, which has provided billions of dollars recently.

And so Ukraine is already using some of that money to buy, uh, many of its own drones that it's producing now and sending them to the front lines for their troops.

Well, given all this confidence that Ukrainians have, are people in Kyiv talking about an end to the war and what that could potentially look like?

Not really.

What was your read on the president's messaging and, and the future of Taiwan [laughs] as this on the vanguard of, of 21st century technology?

Well, I would say, you know, first off, the US has been getting more and more nervous, not just under Trump but over the last decade, about China's capabilities.

So China's wanted to seize Taiwan, as you said, since 1949, but it's only in the last couple of years where Chinese military capabilities have grown to the point where they might now be able to succeed in doing so.

If there was a war right now, nobody knows how it would play out, and that's increased concern in Washington about does China have both the intent and the capability to do something around Taiwan? And that's happened just as Taiwan has gotten more and more important as a partner enabling Silicon Valley to produce all of the chips that we need in data centers, in smartphones, in computers.

And so right now, Taiwan is really at the absolute center of the AI industry, and the US can't do AI without chips from Taiwan, which means that on the one hand, we've got to treat the Taiwan question very carefully to make sure-

Okay.

And I think if you, if you look at, you know, China's foreign policy, we think of just the US-China relationship, but look at the China-Japan relationship.

China's caused it, one of its largest trading partners to turn increasingly anti-China because of the rare earth dispute in 2011 and a series of territorial issues.

In your view, is the promise that many people think AI is supposed to deliver Will it actually be delivered? What's the future that you envision after all this capex is spent?

Well, I think it's in some ways funny when we ask, will AI deliver? Because if you look just three years ago, even after ChatGPT was released, there are so many things that were not possible that are possible today, whether it's the number or the scale of hallucinations and answers that chatbots give you, or the scale of work that you can put together today that just wasn't possible in the past.

We've already had so much new capability generated just in the couple of years since ChatGPT that in some ways I think it's an absurd question to ask, will AI deliver? It already has in a lot of ways.

But I think I understand why there's plenty of questions about what about the investment that's happening right now? Will that investment And I think there are reasons to think carefully about how much each company is putting in over what time horizon.

But I guess when I zoom out, I ask myself, do I want a world in which there's more access to compute or less? And it seems to me that we should, rather than being concerned that there's too much compute being put in the ground, be excited that there's actually companies willing to invest in the infrastructure that's going to deliver all these capabilities.

So my question is, do those companies start to see that? Do we live in a world where this is a layer just like the Internet was a layer of technology?

think there's no doubt that we're going to have ai as a layer that's embedded into all technology that we use and when i look at the economic question i say first off is it profitable to serve ai systems today not train the next generation but serve today's and if you listen to what's publicly reported about open ai or anthropic their margins on their inference business are not just positive, but quite good.

And so that, I think, is pretty strong evidence that the delivery of already existing AI services is a pretty profitable business.

What do you make of this?

Yeah, I mean, I think we know that the Chinese customers, ByteDance, Alibaba, and others, they want the chips.

No doubt about, uh, that.

They don't have access to domestic alternatives at the scale that they need.

The Chinese government though is less certain.

Um, the Chinese government wants to make sure its companies have the computing resources that they need, but China also wants to build up its own domestic ecosystem to support Huawei and the other companies that are designing AI accelerators.

And I think the Chinese government has been, uh, going back and forth on how many GPUs it's going to allow in because it wants to make sure that there's plenty of space in the market for Huawei, because in the long term, the strategy is to supplant foreign s- uh, providers of GPUs, uh, with Huawei's own- Domestic production.

Where do you think it goes on that time horizon?

So, Chris, if you look back on the CHIPS Act, which, of course, was undertaken to really start that kind of investment in our chip industry, to what extent has it reshaped global competition so far? And do you think there have been unintended consequences that have occurred for places like South Korea, Japan, or Taiwan, or might have unintended consequences going forward?

Well, I think it's pretty clear that, again, relative to the prior trend, the CHIPS Act did spur a major increase in investment in chip making facilities.

If you look at the data, there's something like a 10x increase in investment in chip plants over the last couple of years relative to the trend over the prior decades.

That's very clearly driven by the CHIPS Act.

I think it's still the case that the U.S. has a whole range of bottlenecks when it comes to building out advanced manufacturing capabilities, labor, permitting, energy availability, and many others.

But I think actually when you when you zoom out and ask about international cooperation, what you find is that there's been a lot more cooperation than one might have expected.

It's notable that some of the key investments that were supported by the CHIPS Act were from non-U.S. companies.

So how do you think the U.S. should balance export controls with this need to maintain leadership and innovation and in markets and protect our national security?

How are you looking at their capabilities when it comes to the ground up? Because there's a lot of attention being paid to the sector right now.

You know, I think China has made real strides in catching up in many segments of the chip industry, but there are also a number of areas where they're far away from the cutting edge.

If you look, first off, at the manufacturing of semiconductors, uh, what you find is that TSMC in Taiwan is not only able to produce significantly better chips, and we know this because the capabilities of the leading Chinese chip maker, SMIC, basically track along TSMC at a five or six-year delay and that delay has been relatively steady over time.

So there's a meaningful delay in terms of quality and there's a huge gap in terms of quantity because TSMC can access the tools that are used to make chips, uh, machines that are produced by companies like ASML or Applied Materials, in ways that China can't because of export control.

So on- on manufacturing quality and quantity, there's a very big gap.

When it comes to chip design, the gap is much, much less significant which is why Huawei, for example, is a top-notch designer of semiconductors both for smartphones and also increasingly for AI chips.

But design only matters if you can get them manufactured and that's the challenge that Huawei and the other, uh, Chinese chip designers have faced and so I think if you look at the startup ecosystem in China, for example, companies like More Threads, which you just mentioned had an IPO earlier this month, very successful IPO, valuing it at around $50 billion.

How are you seeing this AI broader competition between the US and China and do you think this recent announcement of the H200s really changes the dynamic in this AI competition between the two powers?

Well, lay out the core national security concerns here.

I think there are two main categories of national security concern.

First is that it's already the case that all of the world's leading militaries and intelligence agencies are using AI for their own purposes.

You already see drones powered by AI flying in the skies between Russia and Ukraine.

You already see intelligence agencies using AI to sift through information.

And this is only gonna continue.

And so there's a strong reason to believe that whichever country has the best AI capabilities for its military will have the strongest, uh, military power.

Are either of those arguments persuasive in your view?

TSMC have any kind of programs in place where it sort of tries to help some of these smaller chip makers sort of accelerate their business at all?

It does, and not specifically focused on GPUs, but across the board, there are a number of programs that TSMC uses to help smaller firms build up their capabilities because it is so hard to go from startup to producing chips at scale.

But I think TSMC also has to balance the desire to help new firms with the reality that a small number of big tech firms, Apple, NVIDIA, chief of them constitute a really substantial portion of their revenue.

And so that's naturally where their focus is going to lie.

So how does TSMC help some of those smaller customers grow? Well, if you're a small firm, you've got to manage all sorts of complex IP issues.

You've got to learn how to convert your chip from essentially a software file to a real piece of silicon.

There's a lot of really unique know-how that TSMC has cultivated over the past four decades now that can help young firms learn how to replicate.

How much are these customers thinking about the software that goes on top of these chips from the people that you've talked to?

And I think that's a... think maybe we'll throw this one to chris i mean this is a this is like a like this is this is like a grand strategic question almost of like how much of china's rise is okay and how much of it isn't because um you know the the military intelligence human rights applications are almost secondary to like just how scary you see a richer more flourishing more powerful china to be

right Yeah, and I would even segment out the sort of richer and more flourishing side and just talk about technological capabilities.

They're obviously interlinked, but I don't think U.S. strategy has been to try to make China poorer or less flourishing in general.

I think the question is just who's going to lead in AI? The trend over the last five years and really the last 50 years has been that if you want advanced AI, you need lots of advanced computing.

And there's a small number of companies who produce the chips in question.

Um, and so if you think that advanced technology has mattered in the past in TO streets competition, which I think is pretty hard to argue with, it's probably gonna matter in the future.

And therefore, who wins in AI, you know, it's a little bit reductive to ask, but I think it also, uh, it matters.

Chris, what's your what's your read on this like kind of interesting brushback pitch we've gotten from the central organs about H20s in China?

Do you agree with him? Do you think these measures are in some way self-defeating?

Well, I think there's no doubt that in the short run, it reduces NVIDIA sales.

That's certainly clear.

So does it hurt that business? I think the answer is yes.

One of the dilemmas that any sort of export control regime faces is that if you restrict the sale of a certain tool, you hurt the tool makers, but you benefit the domestic companies that use that tool who have less foreign competition.

So, for example, if you restrict the sale of chip making tools, it's bad for the companies that make chip making tools, but it's actually better for chip makers in the West, including in Taiwan and South Korea, because they don't have strong competition in China at the cutting edge.

If you thought about the counterfactual, suppose the Trump administration had not leaned on the Netherlands in 2019 and said the Netherlands don't sell EUV tools to China, China would be operating today a fleet of many.

How impactful was this to, you know, like the whole AI system?

Mm-hmm.

Huawei is very capable at chip design.

It's done it for smartphones at the cutting edge.

Now it's, you know, catching up very rapidly in terms of GPU design.

Uh, hard to exactly, uh, benchmark vis-a-vis NVIDIA, but Huawei is very capable in chip design.

The, the challenge it had is that SMIC, um, is ab- right now about five years behind TSMC, five or six years, in terms of its ability to produce processor chips at quality and at scale, uh, and so that has been a, a real limiting factor.

And the reason SMIC's behind is because there are no real Chinese equivalents today of companies like ASML or Applied Materials or Tokyo Electron that make the tools that make chips.

... activity, buying from, from TSMC and, and other players? D- Have you tracked that closely?