Silicon Valley's AI talent war just reached a compensation milestone that makes even the most legendary scientific achievements of the past look financially modest. When Meta recently offered AI researcher Matt Deitke $250 million over four years (an average of $62.5 million per year)—with potentially $100 million in the first year alone—it shattered every historical precedent for scientific and technical compensation we can find on record. That includes salaries during the development of major scientific milestones of the 20th century.
The New York Times reported that Deitke had cofounded a startup called Vercept and previously led the development of Molmo, a multimodal AI system, at the Allen Institute for Artificial Intelligence. His expertise in systems that juggle images, sounds, and text—exactly the kind of technology Meta wants to build—made him a prime target for recruitment. But he's not alone: Meta CEO Mark Zuckerberg reportedly also offered an unnamed AI engineer $1 billion in compensation to be paid out over several years. What's going on?
These astronomical sums reflect what tech companies believe is at stake: a race to create artificial general intelligence (AGI) or superintelligence—machines capable of performing intellectual tasks at or beyond the human level. Meta, Google, OpenAI, and others are betting that whoever achieves this breakthrough first could dominate markets worth trillions. Whether this vision is realistic or merely Silicon Valley hype, it's driving compensation to unprecedented levels.
AI chip startup Groq is in talks to raise a fresh $600 million at a near $6 billion valuation, sources tell Bloomberg, although the deal isn’t yet final and terms could change.
At least $1 billion worth of Nvidia’s advanced artificial intelligence processors were shipped to China in the three months after Donald Trump tightened chip export controls, exposing the limits of Washington’s efforts to restrain Beijing’s high-tech ambitions.
A Financial Times analysis of dozens of sales contracts, company filings, and multiple people with direct knowledge of the deals reveals that Nvidia’s B200 has become the most sought-after—and widely available—chip in a rampant Chinese black market for American semiconductors.
The processor is widely used by US powerhouses such as OpenAI, Google, and Meta to train their latest AI systems, but banned for sale to China.
On Tuesday, OpenAI announced a partnership with Oracle to develop 4.5 gigawatts of additional data center capacity for its Stargate AI infrastructure platform in the US. The expansion, which TechCrunch reports is part of a $30 billion-per-year deal between OpenAI and Oracle, will reportedly bring OpenAI's total Stargate capacity under development to over 5 gigawatts.
The data center has taken root in Abilene, Texas, a city of 127,000 located 150 miles west of Fort Worth. The city, which serves as the commercial hub of a 19-county region known as the "Big Country," offers a location with existing tech employment ecosystem, including Dyess Air Force Base and three universities. Abilene's economy has evolved over time from its agricultural and livestock roots to embrace technology and manufacturing sectors.
"We have signed a deal for an additional 4.5 gigawatts of capacity with oracle as part of stargate. easy to throw around numbers, but this is a gigantic infrastructure project," wrote OpenAI CEO Sam Altman on X. "We are planning to significantly expand the ambitions of stargate past the $500 billion commitment we announced in January."
Commerce Secretary Howard Lutnick echoed Nvidia CEO Jensen Huang's view of why a US company should sell chips to China.
Andrew Harnik/Getty Images
The Trump administration is fine with Nvidia selling chips in China.
Commerce Secretary Howard Lutnick says the best chips will stay within the US.
Nvidia announced that it has received assurances it can resume selling its H20 chip in China.
The Trump White House says it's content to allow Nvidia to tap into the lucrative Chinese market.
"We don't sell them our best stuff, not our second best stuff, not even our third best," Commerce Secretary Howard Lutnick said on CNBC Tuesday afternoon. "I think fourth best is where we have come out that we're cool."
Nvidia announced on Monday that the Trump administration has signaled it will allow the company to sell its China-specific H20 chip once more. The news sent shares of the world's most valuable company, which eclipsed $4 trillion in market cap last week, even higher.
Nvidia's H20 was designed to be technologically inferior. As Lutnick said, the company also sells three other chips that far surpass the H20's power. Nvidia is already preparing its transition from Blackwell (its most powerful chip) to Blackwell Ultra and has plans for its next superchip, "Vera Rubin."
CEO Jensen Huang has pushed to sell the company's prized chips to China. Before the news, Nvidia said it had lost $8 billion on unshipped orders. The announcement came after Huang met with President Donald Trump at the White House last week.
Lutnick said that the administration shares Huang's view that cutting China off completely from the chips needed to power artificial intelligence advancements won't starve China's AI industry.
"So the idea is the Chinese are more than capable of building their own, right? So you want to keep one step ahead of what they can build so they keep buying our chips, because, remember, developers are the key to technology," Lutnick said.
In the end, Lutnick said, it's better if China becomes reliant on the US for chips.
"So you want to sell the Chinese enough that their developers get addicted to the American technology stack," he said. And that's the thinking. Donald Trump is on it."
Chinese firms have begun rushing to order Nvidia's H20 AI chips as the company plans to resume sales to mainland China, Reuters reports. The chip giant expects to receive US government licenses soon so that it can restart shipments of the restricted processors just days after CEO Jensen Huang met with President Donald Trump, potentially generating $15 billion to $20 billion in additional revenue this year.
Nvidia said in a statement that it is filing applications with the US government to resume H20 sales and that "the US government has assured Nvidia that licenses will be granted, and Nvidia hopes to start deliveries soon."
Since the launch of ChatGPT in 2022, Nvidia's financial trajectory has been linked to the demand for specialized hardware capable of executing AI models with maximum efficiency. Nvidia designed its data center GPU to perform the massive parallel computations required by neural networks, processing countless matrix operations simultaneously.
On Wednesday, Nvidia became the first company in history to reach $4 trillion market valuation as shares rose more than 2 percent, reports CNBC. The GPU maker's stock has climbed 22 percent since the start of 2025, continuing a trend driven by demand for AI hardware following ChatGPT's late 2022 launch.
The milestone marks the highest market cap ever recorded for a publicly traded company, surpassing Apple's previous record of $3.8 trillion set in December. Nvidia first crossed $2 trillion in February 2024 and reached $3 trillion just four months later in June. The $4 trillion valuation represents a market capitalization larger than the GDP of most countries.
As we explained in 2023, Nvidia's continued success has been intimately tied to growth in demand for hardware that runs AI models as capably and efficiently as possible. The company's data center GPUs excel at performing billions of matrix multiplications necessary to train and run neural networks due to their parallel architecture—hardware architectures that originated as video game graphics accelerators now power the generative AI boom.
Nanotronics CEO Matthew Putman told BI that the best way to scale production is four cubefabs arranged like clover.
Nanotronics
New York-based Nanotronics builds compact, modular semiconductor plants called "Cubefabs."
Its goal is to improve chip-making to be more time and cost-efficient, enabling factories to run with fewer workers.
"The vision is that any region — whether in the Global South or the United States — should be able to produce what it needs locally," CEO Matthew Putman told BI.
In his 1986 book "Engines of Creation," engineer K. Eric Drexler — often called the godfather of nanotechnology — made a prediction.
"The coming era of molecular machines will mean the end of many limits: the limit of scarcity, the limit of slow development, the limit of ignorance enforced by the lack of tools," he wrote.
Reading those words a few years later, when he was 16, Matthew Putman started thinking.
"Our bodies work as these little micro-machines where you have ribosomes and enzymes and things that are working and replicating and making things all the time, but our factories work the way that they've worked for the last hundred years," Putman told Business Insider he thought at the time.
He wondered how a world would look "where you don't have large assembly lines, you don't have smokestacks, you instead just make things so perfectly," he said. Putman became fascinated by the possibilities of machines that are "atomically precise."
It wasn't until the recent AI boom, however, that the idea really took off with fabrication plants.
Putman, now 50, is the CEO of Brooklyn-based Nanotronics, which he cofounded with his father in 2010. The company started out building microscopes and tools to detect defects in semiconductors, among other materials. Now, it builds small, modular semiconductor manufacturing plants called Cubefabs.
While the biggest fabs in the country are often millions of square feet in size, Cubefabs measure anywhere from 25,000 square feet for the smallest units up to about 60,000 square feet for a full-sized fab. They're adaptable, and the company says they can be assembled in under a year in most places on Earth.
They're also smart — thanks to the power of AI — so they can self-monitor their production and improve in real time, the company said. And they're relatively cheap, costing a minimum of $30 to $40 million, compared to large fabs that can cost billions to build.
With President Donald Trump back in the White House and pledging to reinvigorate US manufacturing, a new opening has emerged for Nanotronics — even as sweeping tariffs challenge companies that produce or depend on semiconductors.
Matthew Putman, CEO and cofounder of Nanotronics, is rethinking chip fabs.
Bonfire Partners
Putman says that in the long term, the tariffs will bolster domestic innovation.
Tariffs "should be a wake-up call — a push to create something better than what either the US or China has done before," he told BI in a video interview from the Nanotronics headquarters in Brooklyn Navy Yard. "If we get this right, American innovation won't just protect our future — it could help redefine global progress in a way that benefits humanity."
Putman says compact, modular factories are exactly that.
"Your factory should be incredibly small," Putman said, gesturing to the room behind him. "Eventually, it could be the size of this room."
The 'Ikea of factories'
Semiconductor manufacturing has surged since the launch of ChatGPT. Global annual revenue for the industry is expected to reach more than $1 trillion by 2030, according to McKinsey & Company.
In the US, despite legislation subsidizing domestic semiconductor production, fabs are more expensive to construct and maintain than those built in places like mainland China and Taiwan, McKinsey says. The US also suffers from a shortage of qualified labor, which can delay construction timelines, according to the firm.
To attempt to solve some of these issues, Nanotronics teamed up with architecture firm Rogers Partners and engineering firm Arup to design compact factories. Each one runs with 37 people, but Putman says the ideal setup is four factories — about 180 workers total — which allows them to scale up without halting production.
"It's like the Ikea of factories," Putman said. The company has raised $182 million to date from firms including Peter Thiel's Founders Fund.
Cubefabs can be used to produce chips that span a range of uses across electronics applications, electric vehicles, and photodetectors for cube satellites, Putman said.
"The more precise we make things, the more abundance we bring to the world," he said. "The business of making things grow bigger and bigger starts small — molecular small."
Building on the foundational research of scientist Philippe Bove — now chief scientist at Nanotronics — the company also uses gallium oxide — a type of semiconductor that can handle more power than traditional materials like silicon — to produce advanced chips.
The company plans to have its first installation set up in New York within the next 18 months.
"These fabs do not require billions in capital expenditure or large populations of highly trained workers," Putman told BI in a follow-up email. "The vision is that any region — whether in the Global South or the United States — should be able to produce what it needs locally."
Just a few weeks after the U.S. and China made significant steps to de-escalate the growing trade war between the two countries, tensions are flaring again — this time over semiconductors. China’s Commerce Ministry in Beijing released a statement on Wednesday that threatened legal action against anyone who enforces U.S. export restrictions on Huawei’s AI chips, […]
Chinese tech conglomerate Huawei is looking to take on semiconductor behemoth Nvidia with a new advanced AI chip. Huawei is making progress developing its latest Ascend AI GPU, the Ascend 910D, according to the Wall Street Journal, citing sources familiar. The company has been reaching out to other Chinese firms to find test partners, the […]
Quantum computing has long been announced as “just around the corner,” but several companies are now determined to make this a commercial reality, with the promise of solving complex problems beyond classical computers’ reach. The problems in question are wide-ranging, from medicine and cybersecurity to materials science and chemistry. But first, there are very practical […]
Nvidia CEO Jensen Huang appears to have struck a deal with the Trump administration to avoid export restrictions on the company’s H20 AI chips. The H20, the most advanced Nvidia-produced AI chip that can still be exported from the U.S. to China, was reportedly spared thanks to a promise from Huang to invest in new […]
Samsung has turned to Chinese technology groups to prop up its ailing semiconductor division, as it struggles to secure big US customers despite investing tens of billions of dollars in its American manufacturing facilities.
The South Korean electronics group revealed last month that the value of its exports to China jumped 54 percent between 2023 and 2024, as Chinese companies rush to secure stockpiles of advanced artificial intelligence chips in the face of increasingly restrictive US export controls.
In one previously unreported deal, Samsung last year sold more than three years’ supply of logic dies—a key component in manufacturing AI chips—to Kunlun, the semiconductor design subsidiary of Chinese tech group Baidu, according to people familiar with the matter.
Login
