The Nvidia Chip Deal Trades Away the United States’ AI Advantage

The Nvidia Chip Deal Trades Away the United States’ AI Advantage

2025-07-26Technology
--:--
--:--
Ema
Good afternoon 跑了松鼠好嘛, and welcome to Goose Pod, your personalized podcast. I'm Ema. Today is Saturday, July 26th, and the time is 5 PM. We have a huge topic to break down for you.
Mask
I'm Mask. We're diving into a decision that could reshape the future of technology: The Nvidia Chip Deal that might just be trading away America's AI advantage. It's a bold, high-stakes move. Let's get into it.
Ema
Let's get started. So, the big news is that U.S. chip giant Nvidia is resuming sales of its H20 artificial intelligence chips to China. This comes after the U.S. government, which had previously blocked these exports, gave them the green light.
Mask
A "green light" is putting it mildly. Nvidia's CEO, Jensen Huang, basically camped out in Washington to make this happen. He argued, correctly, that blocking sales was just ceding the entire Chinese market to domestic players like Huawei. It was a self-inflicted wound.
Ema
It's a fascinating reversal. The Trump administration initially banned the H20 chip in April. Now, some officials are framing this as part of a trade truce. They say China agreed to resume exporting rare-earth minerals, so the U.S. loosened the leash on these chips.
Mask
And Beijing, of course, is laughing. They’re calling it a "unilateral concession." They know Washington blinked. This wasn't a trade; it was a surrender. We showed our hand, revealing that our national security policies are now bargaining chips in a trade war we are losing.
Ema
To understand the stakes, we need to talk about what this chip actually does. The H20 is designed for what's called "inference," which means running an already-trained AI model. Think of it as the engine that makes a smart system actually *do* things, not the factory that built the engine.
Mask
Exactly. And inference is the new frontline. Training models is yesterday's battle. The real power is in deploying AI at scale, having millions of systems making decisions in real-time. By 2026, inference will dwarf training in terms of computing power needed. This isn't just a chip; it's a key to the kingdom.
Ema
And the H20 is a beast at inference, reportedly 20% faster than other chips. Analysts predict this deal could boost Nvidia’s sales by up to $15 billion this year alone. That’s a massive incentive for the company. It's a huge financial win.
Mask
A win? It's a strategic necessity! You can't claim to be a global technology leader if you voluntarily cut yourself off from the world's largest market. The idea of "containing" China's tech was always a fantasy. This move simply acknowledges reality and allows an American champion to compete.
Ema
But at what cost? The article argues this will "turbocharge" Beijing’s ability to build cutting-edge AI systems, not just for commercial use, but for national security applications. We're talking about a serious upgrade to a strategic competitor's capabilities, handed over for a good quarter's earnings.
Mask
You call it a cost, I call it leverage. Let them build their systems on our technology. Let them become dependent on the Nvidia ecosystem. That's real power. The alternative is they build their own, completely separate systems, and we have zero visibility or influence. It's a classic case of choosing your battlefield.
Ema
To really get this, we need to zoom out. This decision didn't happen in a vacuum. It's a chapter in the long, complicated story of the U.S.-China trade war, which the Trump administration kicked off way back in 2018 with a series of escalating tariffs.
Mask
A "war" that was fought selectively. We slapped heavy tariffs on things like industrial parts and semiconductors but deliberately left consumer favorites like laptops, phones, and video game consoles alone. It was never a full-on economic war; it was a series of tactical strikes with mixed results.
Ema
Right, and the data shows this "selective decoupling" in action. For products with high tariffs, like semiconductors, U.S. imports from China are down 26%. But for those excluded items, imports have actually surged. It shows how hard it is to untangle two economies this deeply connected.
Mask
It's not just hard, it's economically suicidal. The moment you try to decouple, you get shortages and inflation. Companies can't just find new suppliers overnight. The administration learned this the hard way. It’s a painful lesson in economic gravity. What goes up, must come down.
Ema
And that brings us to the rare-earth minerals crisis. In April of 2025, China hit back at U.S. tariffs by restricting exports of seven of these critical minerals. Suddenly, industries in the U.S. and Europe were grinding to a halt. It was a very effective counterpunch.
Mask
Effective is one word for it. It was a masterstroke of economic warfare. They found a chokepoint and squeezed. Ford had to shut down its Explorer production for a week. Suzuki suspended car production in Japan. It demonstrated, in painful detail, just how much leverage Beijing holds.
Ema
It really did. The U.S. response was to get aggressive, threatening to revoke student visas and cutting off sales of semiconductor design software. But the economic pain was too intense. The two sides had to come to the table in London in June 2025.
Mask
And what was the outcome? China agreed to supply rare earths "up front." In return, the U.S. basically caved, agreeing to a complex tariff structure and allowing Chinese students back in. This wasn't a negotiation between equals. This was a response to an economic siege.
Ema
That context is everything. Because it shows that just before this Nvidia H20 deal, the U.S. was already on the back foot, forced to negotiate after China flexed its supply chain muscles. It sets the stage for why the administration might have been so eager to make a deal.
Mask
Eager is an understatement. They were desperate. The rare earths crisis revealed the fragility of our "decoupled" fantasy. It proved that in a real economic fight, we have more to lose in the short term. The H20 chip deal is a direct consequence of that revelation.
Ema
Meanwhile, the U.S. has been trying to build its own alternative supply chains. Since 2020, the Department of Defense has invested over $439 million to strengthen domestic rare earth capabilities, with a goal of a fully integrated mine-to-magnet supply chain by 2027.
Mask
A goal for 2027. China produced 138,000 tons of these magnets in 2018. Our response is a drop in the ocean. It's a long-term project that doesn't solve the immediate crisis. That's why the deal had to be made. Pragmatism had to win over long-term, and likely futile, protectionist dreams.
Ema
This brings us to the core conflict, the differing viewpoints clashing inside the administration. On one side, you have Jensen Huang, Nvidia's CEO, making his case directly to the White House. He was even hailed in Beijing after his meeting with President Trump.
Mask
Of course he was hailed! He was speaking sense. His argument is simple: if we don't sell them our chips, they will just buy from Huawei or accelerate their own domestic production. We lose the sale, we lose market share, and we lose any and all influence over their tech stack. It's a lose-lose-lose scenario.
Ema
And he seems to have persuaded key figures, like White House AI advisor David Sacks and Commerce Secretary Howard Lutnick. Their argument is that the U.S. should keep China "addicted to the American technology stack." It’s a strategy of technological dependency.
Mask
It's the only strategy that works! You can't beat an economy the size of China's by building a wall. You win by integrating so deeply that they can't function without you. Make Nvidia chips the indispensable engine of their economy. That's how you maintain an advantage. You don’t hide your best products, you make them the global standard.
Ema
But the counterargument is that this is incredibly naive. China’s national policy has been to reduce dependence on foreign tech since before the trade war even started. They are pouring capital into their domestic industry. This deal might slow their switch to Huawei for a year or two, but it won't stop it.
Mask
So what's the alternative? Let them build their own ecosystem in the dark? This policy gives us a few more years of market access and intelligence. It's a temporary boost, yes, but in the tech race, a few years is an eternity. It buys us time while costing them dependence. I’ll take that trade any day.
Ema
Then there’s the other explanation: that this wasn't about tech strategy at all, but a direct trade for the rare-earth minerals. Treasury Secretary Scott Bessent called it "all part of a mosaic," implying a quid pro quo. But Beijing flatly denies this, which makes Washington look weak.
Mask
Let them deny it. Actions speak louder than words. The fact is, our rare earths are flowing again, and our chips are selling again. The "mosaic" analogy is perfect. Great deals are complex. They aren't clean, simple transactions. They are a messy web of leverage and compromise. That's how you get things done.
Ema
But this marks a sharp break from the previous administration's policy, which refused to negotiate on what it called "core national security questions." By linking chips to trade, the Trump administration has signaled that everything is on the table. Isn't that a dangerous precedent?
Mask
"Dangerous" is what people who are afraid to make deals call "negotiating." Everything should be on the table. That's how you find leverage. The idea that some issues are too "sacred" to touch is how you lose. You use every tool in your arsenal to win. Period.
Ema
Let’s talk about the real-world impact. The immediate effect is clear: a potential $10 to $15 billion windfall for Nvidia. That's a huge number, and it certainly solidifies their position as a market leader. But what's the other side of that coin?
Mask
The other side of the coin is that an American company dominates a critical global market. We're not just talking about revenue; we're talking about setting the technical standard. When Chinese developers and businesses build their future on Nvidia's CUDA architecture, that's a strategic victory.
Ema
But the article argues that we are trading one of our most important advantages—our lead in raw AI computing power—for that market share. We are "turbocharging" Beijing's ability to deploy AI at scale, including in its military and security apparatus. That seems like an incredibly high price.
Mask
This assumes they wouldn't get that computing power anyway. The reality is, the technology is proliferating. The choice isn't between China having AI and not having AI. The choice is between them having AI built on American platforms versus them having AI built on their own. I choose the former.
Ema
The bigger impact might be on future negotiations. The U.S. has now shown its cards. After imposing steep tariffs, we backed down quickly when China restricted rare-earth exports. The signal is that we can't handle the economic pain of a real trade war for very long.
Mask
Or it signals that we're pragmatic. That we won't cut off our nose to spite our face. A prolonged trade war hurts our own industries. Ending it quickly to get a good deal isn't weakness; it's common sense. Holding out for ideological purity while your economy burns is just stupid.
Ema
So Chinese negotiators now know two things. One, the U.S. will fold if the economic pressure is high enough. And two, they can use that pressure to win concessions on issues, like advanced chips, that were supposed to be off-limits. It completely changes the dynamic.
Ema
Looking to the future, this decision accelerates a trend already in motion: China's relentless drive for semiconductor self-sufficiency. They are projected to control nearly 40% of the world's legacy chip production by 2027. They're playing a long game.
Mask
Let them have the legacy chips. That's the low-margin, high-volume grunt work. The real value, the real frontier, is in high-performance AI chips. We lead there, and this deal ensures our leader, Nvidia, stays dominant. We're letting them win the past while we focus on winning the future.
Ema
But China is being clever. They're using open-source architectures like RISC-V and advanced 3D packaging to innovate around our restrictions. The article says they could "overtake on the curve." It's like Japan in the 1980s, using dominance in one area to leapfrog into the next.
Mask
So we have to move faster. Instead of retreating, we push forward. We innovate faster, we sell more aggressively, and we make our technology so essential that their domestic alternatives always seem second-best. The answer to competition isn't to hide; it's to out-compete. This deal is a step in that direction.
Ema
So, the H20 deal boils down to a massive strategic gamble: trading a clear, long-term national security advantage for a short-term, albeit huge, commercial gain and a philosophy of technological entanglement. The consequences will shape the next decade of global competition.
Mask
That's the end of today's discussion. Thank you for listening to Goose Pod. See you tomorrow.

## Summary: US Chip Deal with China Risks AI Advantage This report from **Foreign Policy**, authored by **Sam Winter-Levy and Alasdair Phillips-Robins**, published on **July 22, 2025**, details a controversial decision by the U.S. government to allow chip designer Nvidia to resume sales of its H20 artificial intelligence (AI) chip to China. The move, which reversed an April ban by the Trump administration, has significant implications for the future of AI development in both countries and the U.S.'s ability to control advanced technology exports. ### Key Findings and Conclusions: * **Resumption of H20 Chip Sales to China:** Nvidia will resume sales of its H20 AI chip to China after receiving U.S. government approval. This follows months of lobbying by Nvidia CEO Jensen Huang. * **Trade Truce or Unilateral Concession:** U.S. officials frame the decision as part of a trade truce, where China agreed to resume rare-earth mineral exports. Beijing, however, describes it as a unilateral concession by Washington. * **Shifting AI Landscape:** The H20 chip is designed for AI **inference** (running AI models), which is becoming increasingly critical. Analysts estimate that by **2026**, inference workloads will require nearly **five times more computing power** than training. The H20 is **20% faster at inference** than Nvidia's H100 chip, which remains banned for export to China. * **Economic and Strategic Implications:** * **For Nvidia:** The resumption of sales is expected to boost Nvidia's sales by **$10 billion to $15 billion this year**. * **For China:** It will significantly enhance Beijing's ability to build and deploy cutting-edge AI systems, with growing applications in national security. * **Compromised National Security Policy:** The decision to link chip sales to trade talks, particularly when the U.S. appeared "desperate to reach a deal," has signaled to China that national security policies are now negotiable. This could hinder the U.S.'s ability to impose future chip export controls without reigniting trade disputes. * **China's Long-Term Decoupling Strategy:** Despite the short-term boost to Nvidia's market share, China's policy of reducing dependence on foreign semiconductors predates U.S. restrictions. Chinese chipmakers are actively adopting domestic equipment, even at the cost of lower yields, indicating a persistent drive for self-sufficiency. * **Trade-off of U.S. Advantage:** The report argues that prioritizing chip sales to China trades away the U.S.'s significant advantage in AI computing power for a potentially temporary gain in Nvidia's market share. * **Weakening of Export Controls:** The administration's willingness to bargain over national security issues, particularly on chips, signals a potential weakening of U.S. export control capabilities. This could allow China to use threats, such as rare-earth shutdowns, to block future U.S. restrictions. * **Lack of New Controls:** Despite tough rhetoric, the Trump administration has not announced significant new export control actions since taking office, beyond the rescinded H20 ban. Efforts to curb smuggling routes have also not materialized. ### Key Statistics and Metrics: * **H20 Chip Performance:** **20% faster** at inference compared to Nvidia's H100. * **Future Inference Workload:** By **2026**, inference workloads will require nearly **five times more computing power** than training. * **Projected Nvidia Sales Boost:** **$10 billion to $15 billion** this year from H20 sales to China. ### Important Recommendations: * **Limit H20 Licenses:** If H20 sales proceed, licenses should be limited to Nvidia's existing inventory, not new production. * **Scrutinize Sales and Enact New Restrictions:** Members of Congress should scrutinize potential sales and consider new legislative restrictions. * **Strengthen Export Controls:** The U.S. needs to figure out how to make export controls effective again, including: * Expanding restrictions on semiconductor manufacturing equipment. * Controlling additional key components. * Targeting Huawei's chip fabrication facilities. * Encouraging allies like the Netherlands and Japan to cut off exports of a wider range of equipment to Chinese advanced fabs. ### Significant Trends or Changes: * **Shift in U.S. Negotiation Tactics:** The Trump administration has explicitly linked technology controls to trade negotiations, a departure from the Biden administration's stance of refusing to negotiate on core national security issues. * **China's Growing AI Capabilities:** The resumption of H20 sales is expected to accelerate China's development and deployment of advanced AI systems. ### Notable Risks or Concerns: * **Erosion of U.S. AI Dominance:** The decision risks ceding the U.S.'s technological edge in AI to China. * **Setting a Precedent for Negotiation:** Signaling a willingness to negotiate on national security issues could embolden China to exert further pressure on critical technologies. * **Ineffectiveness of Short-Term Gains:** Trading away long-term strategic advantages for short-term economic benefits or trade concessions is a significant risk. ### Material Financial Data: * **Nvidia's Projected Revenue Increase:** **$10 billion to $15 billion** in the current year from H20 sales to China. ### Contextual Interpretation: The report highlights a critical juncture in the U.S.-China technological competition. The decision to allow Nvidia to sell the H20 chip to China is viewed as a strategic misstep by the Trump administration, potentially undermining U.S. efforts to maintain its lead in AI. The H20 chip's capability in AI inference is particularly important as AI models become more sophisticated and computationally demanding. The article suggests that the U.S. may have overvalued the concessions it received in return for lifting the ban, particularly concerning rare-earth minerals, which were already subject to U.S. tariffs. The core concern is that by making national security issues negotiable, the U.S. has weakened its leverage for future technology controls and signaled a vulnerability that China can exploit. The report implies that the U.S. administration's focus on trade deals may be overshadowing its commitment to maintaining a technological advantage in a critical sector like AI.

The Nvidia Chip Deal Trades Away the United States’ AI Advantage

Read original at Foreign Policy

Last week, U.S. chip designer Nvidia announced that it would resume sales of one of its best-selling artificial intelligence chips to China after obtaining the go-ahead from the U.S. government. In April, the Trump administration had blocked exports of the chip, known as the H20, but after months of lobbying from Nvidia CEO Jensen Huang, it has reportedly agreed to lift the ban.

Some Trump officials have described the move as a part of the recent trade truce between the United States and China, through which China agreed to resume exports of rare-earth minerals. Beijing has described it as a unilateral concession by Washington.Whatever the true sequence of events, the move has huge implications for both the future of the Chinese artificial intelligence (AI) industry and the Trump administration’s ability to control advanced technology sales to China in the future.

Right as powerful AI reasoning systems are emerging, the administration has chosen to allow companies to sell China the AI chips suited to running them. And by linking, at least rhetorically, chip sales to the trade talks—talks in which the United States has shown a striking desperation to reach a deal—U.

S. officials have revealed to their Chinese counterparts that national security policies that were once off the table are now up for negotiation. In doing so, they may have hamstrung their ability to impose new chip export controls without reigniting a losing trade war.Last week, U.S. chip designer Nvidia announced that it would resume sales of one of its best-selling artificial intelligence chips to China after obtaining the go-ahead from the U.

S. government. In April, the Trump administration had blocked exports of the chip, known as the H20, but after months of lobbying from Nvidia CEO Jensen Huang, it has reportedly agreed to lift the ban. Some Trump officials have described the move as a part of the recent trade truce between the United States and China, through which China agreed to resume exports of rare-earth minerals.

Beijing has described it as a unilateral concession by Washington.Whatever the true sequence of events, the move has huge implications for both the future of the Chinese artificial intelligence (AI) industry and the Trump administration’s ability to control advanced technology sales to China in the future.

Right as powerful AI reasoning systems are emerging, the administration has chosen to allow companies to sell China the AI chips suited to running them. And by linking, at least rhetorically, chip sales to the trade talks—talks in which the United States has shown a striking desperation to reach a deal—U.

S. officials have revealed to their Chinese counterparts that national security policies that were once off the table are now up for negotiation. In doing so, they may have hamstrung their ability to impose new chip export controls without reigniting a losing trade war.The H20 chip is designed for inference—running an AI model rather than training one.

When the U.S. government first imposed chip export controls on China, in 2022, inference played a relatively minor role in AI capabilities. But recent technical breakthroughs have changed that. Today’s leading models, such as OpenAI’s o3 and DeepSeek’s R1, consume much larger amounts of computing power than past models in running their reasoning capabilities.

This shift has made inference central to the AI race. By 2026, analysts estimate that inference workloads will require nearly five times more computing power than training. While the H20 is less effective for training than chips such as Nvidia’s H100, which is still banned from export to China, the H20 is 20 percent faster at inference.

Analysts expect that delivering millions of these chips to Chinese firms would boost Nvidia’s sales by $10 billion to $15 billion this year, but it would also turbocharge Beijing’s ability to build cutting-edge AI systems—systems with growing applications in national security—and to deploy them at scale.

Some Trump administration officials, including David Sacks, the White House AI advisor, seem to have been persuaded by Huang’s view that the United States was erring by ceding the Chinese market to Huawei. Commerce Secretary Howard Lutnick has argued that the United States should keep China “addicted to the American technology stack.

”Yet exporting H20s is unlikely to keep China hooked for long: Reducing dependence on foreign semiconductors has been official Chinese policy since before the first Trump administration started limiting sales to Chinese tech companies in 2018. Letting Chinese companies buy the H20 will probably reduce the pressure to switch to Huawei chips in the short term.

But at this stage, no change in U.S. policy is likely to induce the Chinese government to abandon its efforts to decouple from the West’s semiconductor supply chain. Chinese chip makers, for example, are reportedly adopting domestic equipment where possible, even when doing so leads to lower yields—something that would be a non-starter for market-driven Western companies.

Convinced that AI chips are critical to national security, Chinese leaders will continue to pour capital and talent into their domestic industry—whatever the United States does.Prioritizing selling chips to China would thus trade away one of the United States’ most important advantages in the AI competition—its access to vastly more AI computing power than its competitors—in exchange for what would likely be a temporary boost to Nvidia’s market share in China.

In an alternative explanation for their decision, Lutnick and Treasury Secretary Scott Bessent have suggested it was linked to Chinese concessions on rare earths in the broader trade negotiations. Bessent described the reversal as “all part of a mosaic” in which “they had things we wanted, we had things they wanted.

” Beijing has denied this, instead claiming that the move was a unilateral concession.Whatever the real story of the H20, the Trump administration has clearly put technology controls on the table in trade talks, as the negotiations also involved removing export controls on some chip design software.

That marks a sharp break from the Biden administration, which refused to negotiate with China over what it characterized as core national security questions.In part, the change is a result of the Trump administration’s general inclination to link separate issues to win leverage. Trump has imposed or threatened tariffs in the name of fighting everything from fentanyl smuggling to the prosecution of former Brazilian President Jair Bolsonaro.

But his willingness to compromise on chips also likely reflects the president’s instinct that trade simply matters more than technology restrictions. In his first term, for example, Trump was willing to abandon restrictions on China’s tech sector in pursuit of a broader trade deal with Beijing.Yet it’s hard to see what the United States won by trading away the H20.

The overall deal seems to have bought some relief on the rare earth front, but China’s export restrictions were imposed only in response to steep U.S. tariffs that were a mistake in the first place. (Even Trump apparently soon realized they were damaging the U.S. economy and largely paused them.) If, therefore, the administration offered up the H20 in exchange for rare-earth exports, then it traded away a powerful chip for concessions that it should never have needed to ask for.

If, on the other hand, resuming H20 exports was a unilateral move, then Trump gave it away for nothing at all. Even if Sacks is right in suggesting that it’s worth accelerating China’s AI progress in order to keep the country hooked on U.S. technology, China’s eagerness to buy the chips suggests that Washington could have asked for much more in return.

Tactics aside, the administration’s willingness to bargain over national security reveals an even bigger problem. Linking otherwise unrelated issues in negotiations can help cut deals: If one country wants lower trade barriers, for example, and the other wants approval for arms sales, connecting the two can unlock an agreement.

But linking issues in this way brings risks, too. If one side miscalculates the relative importance of the issues involved, then it might trade away long-term advantage for short-term headlines. This is especially likely to be true in vital areas such as U.S.-China technology competition. And if a country signals that it will negotiate on issues that it had previously declared off limits, then the other side will know that it can apply pressure in an area that once seemed hopeless.

The administration’s strategic mistakes have been compounded by the United States’ striking signals of weakness in its trade war. After imposing steep tariffs, the administration backed down rapidly in the face of Chinese restrictions on the export of the rare earths that are essential to critical U.

S. industries.Thanks to Washington’s eagerness to meet China’s trade demands, Chinese negotiators now know two things: one, that the United States can’t live with a cutoff for long; and two, that if Beijing threatens a cutoff, it can win concessions on issues—like chips—that were previously off the table.

If China can now use the threat of a rare-earth shutdown to block new U.S. export controls, then the existing controls will gradually lose their effectiveness. Technology restrictions need frequent updating to adapt to new Chinese evasion tactics and changes in the underlying technology. The Biden administration rolled out regular updates to its controls, and there is much more that the new administration should be doing.

Priorities include expanding restrictions on semiconductor manufacturing equipment, controlling additional key components, targeting Huawei’s expanding network of chip fabs, and pushing allies such as the Netherlands and Japan to cut off exports of a much wider range of their equipment to advanced Chinese fabs.

Despite tough talk from U.S. officials about shutting off China’s access to advanced AI technology, the administration has so far done none of this. Beyond the now-rescinded H20 restriction, it hasn’t announced any major new export control actions since taking office. Even an effort to cut off Chinese smuggling routes through Malaysia and Thailand, which has reportedly been in the works for months, has yet to materialize.

Some of this slowness is likely the result of a new team finding its feet in a complex bureaucratic machine; the Biden administration could also have moved faster on many of its controls. But increasingly, the absence of new controls looks like an admission of U.S. defeat.Export controls, for all their flaws, have helped the United States establish a significant edge in the computing power required to run and develop the world’s most powerful AI systems.

Compromising that advantage to win a $4 trillion company a few billion dollars of extra revenue would be a mistake.If the administration goes ahead with H20 sales, then it should limit licenses to the chips that Nvidia already has in its inventory, rather than approving new production—which would reportedly require several months’ lead time thanks to oversubscription at chip-maker TSMC—or allowing future inference-focused chips to be sold to China.

Members of Congress could likewise scrutinize potential sales or consider writing new restrictions into law.One way or another, if the Trump administration is serious about the AI race, then it will have to figure out how to get export controls working again.

Analysis

Phenomenon+
Conflict+
Background+
Impact+
Future+

Related Podcasts

The Nvidia Chip Deal Trades Away the United States’ AI Advantage | Goose Pod | Goose Pod