On October 28, a mysterious Silicon Valley startup named Substrate emerged with a bold claim: it has developed a photolithography technology capable of challenging industry giants ASML and TSMC. Founded just three years ago, the company has closed a $100 million seed round, valuing it at over $1 billion. Backers include Peter Thiel’s Founders Fund, General Catalyst, Valor Equity Partners, and In-Q-Tel, a non-profit backed by the U.S. Central Intelligence Agency. The funds will support Substrate’s mission to create a new chip manufacturing method that could cut advanced process costs in half and help rebuild a full-scale semiconductor fabrication ecosystem in the United States. The man behind the vision is 34-year-old British-American James Proud, who has no college degree and no prior experience in semiconductors. Yet he is now aiming to disrupt one of the world’s most complex and tightly controlled industries. In March, Proud sat in the office of U.S. Vice President JD Vance, explaining how his company uses particle accelerators to develop an alternative semiconductor fabrication process. When asked when the technology would be ready, Proud replied simply: “I can only say I made good use of my time during the pandemic.” Proud’s entrepreneurial journey has been unconventional. In 2011, at just 17, he became one of the first recipients of the Thiel Fellowship, a program that provides $100,000 to young innovators who forgo or skip college. His first venture, GigLocator, was a concert information aggregator that he sold in 2012. His second, Hello, launched in 2012, produced a sleep-tracking device called Sense—a tennis-ball-sized sphere that monitored bedroom conditions. The company raised $2.4 million on Kickstarter and attracted $40 million in follow-on funding from investors including Singapore’s Temasek Holdings. At its peak, Hello was valued at $250 million to $300 million. But after five years of struggling with supply chain logistics and retail distribution, the company shut down in June 2017, unable to pay its bills. Attempts to sell to Fitbit failed. Rather than retreat, Proud rebounded. In 2023, he launched Config, a collaboration platform for hardware teams, aiming to solve the coordination problems he encountered during Hello’s development. But his real turning point came when he began studying the geopolitical risks of global semiconductor supply chains around 2019. That year, he became a U.S. citizen and renounced his British nationality. He also drafted policy proposals for President Trump’s first administration, calling for a “Manhattan Project” to revive American chip manufacturing. Today, the semiconductor industry relies almost entirely on EUV (Extreme Ultraviolet) lithography, a technology dominated by ASML. The company’s EUV machines are massive, car-sized systems that use high-powered lasers to vaporize molten tin droplets into EUV light, etching nanoscale circuits onto silicon wafers. Each machine costs about $4 billion and contains over 100,000 components. ASML holds a 100% market share with no real competitors. Substrate’s approach diverges sharply: it’s pursuing X-ray lithography. By generating shorter-wavelength X-rays via particle accelerators, the company aims to project patterns onto silicon with precision comparable to ASML’s most advanced High NA EUV systems—achieving a 12-nanometer feature resolution. If successful, the equipment would be smaller, cheaper, and more adaptable than current EUV machines. So why does Substrate believe it can succeed? Proud has assembled a team of about 50 people, including scientists from Lawrence Livermore National Laboratory and engineers from TSMC, IBM, AMD, Apple, Applied Materials, and Qualcomm. The company claims its technology is entirely proprietary, with no reliance on external intellectual property or equipment. A key differentiator is its use of AI to accelerate R&D. Substrate employs GPUs and TPUs to build end-to-end physics simulations—covering everything from particle accelerators and optics to final transistor design—compressing development timelines from years to days. “We’re not just building a machine,” the company says. “We’re building a new way to design and test it.” In early 2024, Substrate conducted a critical test at a particle accelerator facility in the Bay Area. Initial results were poor—vibrations from the air conditioning system caused image blur. After a full day of troubleshooting, the team adjusted the fan speed, and the system finally produced “very fine and tiny” patterns on a silicon wafer, according to Proud. John E. Kelly, a former IBM semiconductor manufacturing pioneer, reviewed the results and called them “incredibly sharp and clear,” confirming the underlying physics was viable. Still, experts caution that scaling remains a massive challenge. “He’s at the first base camp,” Kelly said, comparing the journey to climbing Mount Everest. “The summit is still far away.” Substrate must now build a $10 billion factory, secure customers, and navigate complex technical and supply chain hurdles. The company’s ambition extends beyond just making a new type of photolithography tool. Its ultimate goal is full vertical integration—designing and manufacturing everything from equipment to wafers in-house. This contrasts sharply with today’s specialized ecosystem, where ASML builds machines, TSMC manufactures chips, and companies like NVIDIA design them. Proud argues this integration reduces costs by eliminating reliance on external suppliers and avoids the need for multiple patterning, a major cost driver in advanced nodes. But critics remain skeptical. Bernstein analyst David Dai notes that X-ray lithography has been explored before and abandoned—ASML itself tested it and gave up. Moreover, Substrate’s strategy eliminates any possibility of forming an industry-wide ecosystem. Experts question whether Substrate can maintain precision across large 300mm wafers at high throughput—something ASML spent decades mastering. Thermal control, vibration isolation, and optical stability are just a few of the engineering barriers that remain. Despite the skepticism, Substrate has attracted serious attention from U.S. government leaders. The Trump administration has prioritized semiconductor production as a national security issue, even discussing purchases of Intel shares. Commerce Secretary Howard Lutnick has met with Proud multiple times. Substrate has also engaged with the Department of Commerce and Department of Energy. However, the Biden administration initially expressed caution. According to two sources, officials worried about Substrate’s plan to use a single particle accelerator to power multiple lithography tools—potentially creating a single point of failure. This led to resistance against a $1 billion request from the CHIPS Act. Proud insists the application was never formally denied and that the design includes redundancy to prevent downtime. Stephen Streiffer, director of Oak Ridge National Laboratory, remains optimistic. “This is a project of national importance. They know what they’re doing,” he said. Proud is driven by a deep sense of mission. He believes the U.S. once led in chip technology and let it slip away too easily. “This is ideological,” he says. “There are plenty of ways to make money in tech, but few choose the hardest path.” Substrate is now selecting a site for its first factory, with discussions underway with Texas A&M University to build a particle accelerator and fabrication facility on campus, estimated at $10 billion. The company has also held preliminary talks with Intel, though the company declined to comment. Proud says the equipment could be deployed in U.S. fabs within “a few years,” with the goal of producing its first chips by 2028. Industry experts consider this timeline overly optimistic. Even if the technology works, Substrate must build an entire manufacturing ecosystem—chemical vapor deposition, ion implantation, polishing, testing, packaging—each with its own barriers and monopolies. Yet for investors like General Catalyst’s Paul Kwan, the risk is worth it. “The only way to break the deadlock is to start from a blank slate,” he said. “Only then can you beat what seems impossible.” Substrate’s story reflects a broader American anxiety and ambition in the semiconductor space. From a non-industry founder backed by Thiel and intelligence agencies, to government leaders seeking alternatives to global monopolies, the push to challenge ASML and TSMC is no longer just technical—it’s geopolitical. X-ray lithography isn’t new. Scientists have tried it for decades. But in an era of supply chain fragility and intense competition, the path once abandoned is now being reconsidered. Other startups like xLight in Silicon Valley and research teams in Japan are exploring similar concepts. The odds of success remain slim. The semiconductor industry’s barriers aren’t just technical—they’re decades of accumulated expertise, massive capital, global supply chains, and institutional trust. ASML’s EUV technology took 20 years and billions to mature. TSMC’s dominance is built on continuous investment. But Proud insists his lack of industry background is an advantage. “If I came from the industry, I’d know how impossible this seems,” he says. “I wouldn’t believe it was possible.” For those betting on Substrate, the message is clear: some rules need to be broken.