Weekly: Who are Jensen's friends at the White House?
8.5 min read.
Highlights
Nvidia not out of the woods. A compelling WSJ article that argues that, although Nvidia has won major victories with the latest permission to sell H20 chips and their breaking the US$4 trillion market cap, Nvidia will still be exposed to the geopolitical winds. If it is true, as Commerce Sec. Howard Lutnick says, that the H20 deal was in exchange for Chinese rare earths, then the Trump admin may turn off the H20 faucet again if the flow of rare earths stop. Plus, there’s the tariffs on semiconductors that are expected to be announced later this month.
Jensen’s friends at the White House. A long NYT piece is a nice complement to the WSJ piece. It seems Jensen Huang has found a friend in David Sacks, Trump’s AI czar. Huang and Sacks have advocated for the sale of AI chips to China to prevent Huawei and other domestic Chinese firms from gaining market share. I’ve written previously about my thoughts here.
Reading between the lines, although David Sacks and Howard Lutnick are both in the pro-Nvidia chip sales camp, there seems to be a subtle distinction, per the NYT piece. Sacks, a Silicon Valley native and friendlier with Jensen, seems to be approaching this as a business / market issue, and it sounds like Sacks would have advocated for H20 sales regardless of the rare earths trade deal.
Lutnick, on the other hand, seems to consider the AI chips as part of a broader geopolitical and commercial strategy on trade with China, and sounds like he would turn off the flow of H20 if the rare earths stopped coming. I am curious if insiders know who has more sway over the president on this issue. It would be interesting to see how Washington reacts to any changes in the trade relationship with Beijing.
How China invests in AI. Another long NYT piece on how China has been investing in AI for the past decade, including building out its manufacturing capabilities, financing data centres and semiconductors, funding a network of labs for AI research, investing in AI startups, hosting neighbourhoods of start-up incubators, creating data resources and lots more. The huge effort doesn’t seem to be super efficient but it’s fast and it’s working.
Table of Contents
Asa Fitch and Dan Gallagher, “Nvidia’s China Troubles Aren’t Over,” WSJ, 07/15/2025.
Tripp Mickle, “How Nvidia’s Jensen Huang Persuaded Trump to Sell A.I. Chips to China,” NYT, 07/17/2025.
Meaghan Tobin, “China Is Spending Billions to Become an A.I. Superpower,” NYT, 07/16/2025.
Jacob Feldgoise and Hanna Dohmen, “Inside Beijing’s Chipmaking Offensive: Where Is China Gaining Ground?,” CST, 07/14/2025.
1.
Asa Fitch and Dan Gallagher, “Nvidia’s China Troubles Aren’t Over,” WSJ, 07/15/2025.
Nvidia notched a political win Monday with the Trump administration’s loosening of restrictions on its chip sales in China. But the artificial-intelligence chip giant isn’t done being a pawn in the big-power rivalry between the U.S. and China.
The reprieve came as boiling trade tensions seem to have reduced to a simmer. The countries reached a trade truce in Geneva in May, and China followed up by easing supplies of much-needed rare-earth metals for U.S. automakers and manufacturers.
But Nvidia is hardly out of the woods. Given its position at the red-hot center of the AI boom, there is no telling which way the political winds might move it next. The Trump administration’s negotiations have been nothing if not volatile. As for China, its rare-earth reprieve only lasts six months. Should that lapse, AI processors could again become bargaining chips for the U.S.
On top of that, there is a tariff threat to keep an eye on. The Trump administration appears poised to impose tariffs on semiconductors, potentially at the 25% rate the president has discussed in the past. Nvidia’s chips, which are almost wholly produced in Taiwan, could be affected.
For now, it is Nvidia’s time to celebrate. The moment to do so could be fleeting.
2.
Tripp Mickle, “How Nvidia’s Jensen Huang Persuaded Trump to Sell A.I. Chips to China,” NYT, 07/17/2025.
In April, Jensen Huang, the chief executive of the chip maker Nvidia, received a blunt welcome to the world of geopolitics when the Trump administration shut down sales of an artificial intelligence chip the company had designed specifically for China.
Since then, Mr. Huang has turned himself into a globe-trotting negotiator as he has tried to persuade President Trump to reverse course. He has traveled with Mr. Trump, testified before Congress and charmed reporters in Washington. And he has courted allies in White House who have quietly supported global business interests despite Mr. Trump’s tough talk on trade with China.
That work has started to pay off for Nvidia. Last week, Mr. Huang met with Mr. Trump in the Oval Office and pressed his case for restarting sales of his specialized chips, said two people familiar with the meeting, who spoke on the condition of anonymity. He argued that American chips should be the global standard and that the United States was making a grave mistake by ceding the giant Chinese market to homegrown rivals.
Mr. Huang, 62, was a reluctant lobbyist. An electrical engineer by training, he used to consider government affairs trivial, said two former employees, who asked to be anonymous because they still work in the tech industry.
David Sacks, a longtime Silicon Valley investor who had become the White House’s A.I. czar, was more receptive to Mr. Huang’s position about China than others in the administration, two people familiar with his thinking said. Mr. Sacks disliked another Biden administration rule that controlled A.I. chip sales around the world. He also questioned Washington’s consensus that selling A.I. chips abroad would be bad for the United States.
In the weeks that followed, Mr. Sacks helped dismantle the Biden rule that put caps on the number of chips Nvidia could sell to every country in the world, four people familiar with the process said. The maneuver cleared the way for Mr. Sacks to help Nvidia sell chips to Saudi Arabia and the United Arab Emirates.
During the sales process, Mr. Sacks and Mr. Huang began advancing the same rationale for selling A.I. chips, two people said. To win the A.I. race, they said, the U.S. government should encourage purchases of U.S. technology rather than create a reason for countries to buy similar Chinese technology.
Mr. Huang delivered that same message to Mr. Trump last week in the Oval Office, two people familiar with the meeting said. Mr. Sacks was seated nearby, lending his support. By the end of the nearly hourlong meeting, Mr. Trump said Nvidia’s chips could return to China.
Mr. Lutnick said on CNBC that the approval was linked to ongoing trade talks with China, which recently agreed to supply rare earth magnets to American companies. The idea was to sell Chinese businesses Nvidia’s fourth-best chip, he said, so that “they get addicted to the American technology stack.”
Days later, Mr. Huang traveled to Beijing and held a news conference to tell customers that Nvidia was open for business. Mr. Huang, now the world’s sixth-wealthiest man, amiably chatted with reporters about his relationship with Mr. Trump. The atmosphere was jubilant.
3.
Meaghan Tobin, “China Is Spending Billions to Become an A.I. Superpower,” NYT, 07/16/2025.
China is quickly closing the gap with the United States in the contest to make technologies that rival the human brain. This is not an accident. The Chinese government has spent a decade funneling resources toward becoming an A.I. superpower, using the same strategy it used to dominate the electric vehicle and solar power industries.
For the past 10 years, Beijing has pushed Chinese companies to build manufacturing capabilities in high-tech industries for which the country previously depended on imports. That approach helped China become the maker of a third of the world’s manufactured goods and a leader in electric vehicles, batteries and solar panels. And it has also been applied to the essential building blocks of advanced A.I. systems: computing power, skilled engineers and data resources.
China pushed that industrial policy approach as three presidential administrations in Washington tried to hold back its ability to make technologies like artificial intelligence, including by restricting sales of chips made by Nvidia, America’s leading A.I. chipmaker.
In the United States, companies like Google and Meta have spent billions on data centers. But in China, it is the government that has played a major role in financing A.I. infrastructure and hardware, including data centers, high-capacity servers and semiconductors.
To concentrate the country’s engineering talent, the Chinese government also financed a network of labs where much of its most advanced A.I. research takes place, often in collaboration with big tech companies like Alibaba and ByteDance.
Beijing has also directed banks and local governments to go on a lending spree that fueled hundreds of start-ups. Since 2014, the government has spent nearly $100 billion on a fund to grow the semiconductor industry, and in April said it would allocate $8.5 billion for young A.I. start-ups.
Local governments have set up entire neighborhoods that function as start-up incubators, like Dream Town in Hangzhou, a city in China’s south that is home to Alibaba and DeepSeek and is known as a hot spot for A.I. talent.
American A.I. systems were built using information from all types of websites, including some that are inaccessible on China’s censored internet, like Reddit and Wikipedia. But Chinese companies need to make sure that any A.I. products used by the general public comply with Beijing’s controls on information. So the government has created data resources that contain approved information for companies to use to train their A.I. systems, like one based on state media articles that is called “the mainstream values corpus.”
Yet Beijing’s industrial policy approach to A.I. has also been inefficient. An abundance of A.I. start-ups are vying for their piece of a cutthroat market, competing to offer their models at low rates to engineers.
This top-down approach also makes it burdensome to shift resources quickly as technology changes. Chinese companies spent years working on A.I. technologies like facial recognition but were caught off-guard by the advances in generative A.I. behind ChatGPT.
4.
Jacob Feldgoise and Hanna Dohmen, “Inside Beijing’s Chipmaking Offensive: Where Is China Gaining Ground?,” CST, 07/14/2025.
Chinese semiconductor manufacturing equipment (SME) companies are gaining market share in several notable technologies, consequently eroding foreign incumbents’ market shares. From 2019 to 2024, Chinese SME firms have made notable gains in equipment used in fabrication—the first stage of chip manufacturing, particularly in chemical mechanical planarization (CMP) tools, dry etch and clean tools, and deposition tools. However, lithography, the most complex tool category, remains a critical constraint for China.
CMP: Hwatsing Technology helped increase China’s market share by eight percentage points (pp) over the five-year period.
Dry etch and clean: China gained parity with Japan in the dry stripping tool market and achieved an eight pp rise in the dry etch market.
Deposition: NAURA, Piotech, and AMEC played a key role in increasing China’s market share in deposition from 2% to 7% over the five-year period.
Lithography: ASML and Nikon continue to control the market, and China’s Shanghai Micro Electronics Equipment (SMEE) has thus far only reached a 4% market share in i-line lithography tools, which are used for legacy chip fabrication.
In assembly, test, and packaging, China is also gaining ground. China’s market share gains in advanced packaging tools may help position the country as an important future player in this stage of the production process, which is becoming increasingly important to AI chip innovations. China’s market share gains in test tools underscores China’s ability to gain a foothold in segments with modest technical barriers.
Advanced packaging: Most advanced packaging tools are fabrication tools specifically designed for the advanced packaging stage of the production process. Over the last five years, Chinese firms’ most pronounced gains in advanced packaging market share occurred in the same equipment classes that propelled their growth in fabrication tools.
Test: Since 2019, Hangzhou Changchuan and AccoTEST have driven China’s market share to almost 70% in linear and discrete test tools and have contributed to solid gains in burn-in and system-on-a-chip (SoC) test tools, while U.S. suppliers ceded ground.
-