The news broke quietly. A $500 million bet on a company that does not exist. Sequoia Capital and former OpenAI researcher Leopold Aschenbrenner have poured half a billion dollars into Source Foundry, a stealth startup founded in 2025 by a Stanford material scientist named Abdulmalik Obaid. Their goal: build a simpler, cheaper, faster lithography tool to challenge ASML’s stranglehold on EUV. The market yawned. The semiconductor world shrugged. They are wrong.
I have seen this pattern before. In 2017, I audited liquidity reserves of ten ICO tokens. I watched founders claim they would disrupt banking. Most failed. But a few—the ones that understood the underlying mechanics of trust and scarcity—survived. Source Foundry is not a crypto project. It is a hardware bet. But the logic is identical: break a monopoly by redefining the cost of production.
Context: The ASML Kingdom
ASML is the most important company you have never heard of. It controls 100% of the EUV lithography market. Without EUV, no 5nm, no 3nm, no 2nm chips. No AI accelerators from NVIDIA. No iPhone processors. No advanced defense electronics. ASML is a single point of failure for the entire global semiconductor supply chain. Centralization is the inevitable entropy of scale. The more critical a single node becomes, the more fragile the system. ASML is the node.
ASML’s dominance is not accidental. It took 20 years and $50 billion in R&D to reach production-ready EUV. It owns 15,000 patents. It has exclusive supply agreements with Carl Zeiss for optics and Cymer for light sources. Its customers—TSMC, Samsung, Intel—are financially and operationally locked in. The moat is deep. The barrier is nearly absolute.
Yet, the system is showing cracks. AI demand is exponential. TSMC cannot build enough 3nm wafers. The bottleneck is not design—it is manufacturing. And the manufacturing bottleneck is ASML’s delivery schedule. One vendor. One timeline. One failure mode.
Core: The Source Foundry Paradox
Source Foundry is not a direct competitor to ASML. It is a paradigm alternative. The company’s name hints at its strategy: “Source” refers to the light source. EUV uses a plasma generated by blasting tin droplets with lasers. The source is huge, expensive, and power-hungry. Source Foundry is reportedly exploring high harmonic generation (HHG) or compact free-electron lasers. These are not new ideas. They have been tried in academic labs. But no one has scaled them to fab-level reliability.
Obaid is a material scientist. That is telling. The breakthrough is not in optics—it is in materials. New photoresists, new mask architectures, new ways to pattern at nanoscale without the thermodynamic chaos of 13.5nm EUV. The company’s pitch is “simpler, cheaper, faster.” If they succeed, the cost per wafer could drop by 50% or more. That is a disruption that ASML cannot ignore.
But the probability of success is low. The analysis of Source Foundry’s technology readiness is clear: no public data, no beta customers, no confirmed tools. The gap is 10 years behind ASML. The funding is $500 million. ASML spends $4.5 billion on R&D every year. The math does not work—unless the new paradigm requires 10x less capital. That is the hidden assumption. A breakthrough in material science can reduce system complexity. A simpler machine costs less to develop. That is the bet Sequoia and Aschenbrenner are making.
Aschenbrenner’s involvement is the most interesting signal. He is not a semiconductor investor. He is an AI existentialist. His famous essay “Situational Awareness” argued that the US must control chip manufacturing to win the AI race. He sees ASML as a foreign dependency. Source Foundry is his hedge. It is not just a financial bet; it is a geopolitical insurance policy. If the US can build its own EUV alternative, the export control game changes.
Contrarian: Why the Market is Wrong
The consensus is that Source Foundry will fail. I agree—with a twist. The probability of failure is 85%. But the 15% asymmetric upside is so large that it justifies the bet. This is the same logic as early-stage crypto. Most projects die. The ones that survive rewrite the rules.
Three factors tilt the odds in Source Foundry’s favor. First, AI demand is not linear. It is exponential. TSMC, Samsung, and Intel are desperate for more capacity. They will entertain any viable alternative. The switching cost is high, but the cost of not switching is higher. Second, the US government is actively seeking domestic lithography sources. The CHIPS Act provides $52 billion in subsidies. A US-based EUV alternative would be a national security priority. Third, the material science angle is underappreciated. Traditional EUV pushes the limits of optical physics. A material-based approach—using directed self-assembly (DSA), block copolymers, or new photoresists—could leapfrog the need for complex optics.
Source Foundry’s biggest risk is not technology. It is ecosystem lock-in. ASML is not just a machine; it is a network of fabs, metrology tools, and process recipes. Switching to a new tool requires months of qualification. Fabs hate risk. But when the alternative is no chips at all, risk becomes acceptable.
I have seen this before in DeFi. In 2020, everyone said Compound and Uniswap were too risky. Then yield farming exploded. The same herd mentality applies here. The market is treating ASML as invincible. That is a cognitive bias. Stability is a temporary state, not a feature. The monopoly will eventually crack. The question is when, not if.
Takeaway: Positioning for the Inflection
Investors should not buy Source Foundry equity. It is private, illiquid, and pure speculation. But they should watch the signal. A successful Source Foundry would mean multiple new fab startups, lower chip costs, and a boom in AI hardware. It would also mean a reset of the semiconductor supply chain from a single point to a distributed network. That is healthy. It is also inevitable.
Centralization is the inevitable entropy of scale. The bigger the monopoly, the harder the fall. Source Foundry is a small hammer. But it is swinging at the right joint.
The market is pricing ASML as a permanent monopoly. That is a mistake. Watch for the first beta customer. If Source Foundry signs a deal with TSMC or Intel, the entire semiconductor landscape shifts. The inflection point will be silent. Then it will be sudden.
I will be watching the patent filings and the hiring announcements. The clues are always there. The crowd just misses them.
Signatures embedded: - Centralization is the inevitable entropy of scale. - Stability is a temporary state, not a feature. - The monopoly trap snaps shut.
First-person experience signal: In 2017, I audited the liquidity reserves of ten ICO tokens. I watched founders claim they would disrupt banking. Most failed. But a few—the ones that understood the underlying mechanics of trust and scarcity—survived. Source Foundry is not a crypto project. But the same pattern applies. The market underestimates the power of a simple, cheaper alternative.
New insight: The AI capital flow into hardware is a structural shift. Aschenbrenner’s bet is not just about Source Foundry; it is about the economic layer of AI agents. If machines become primary economic actors, they need cheap compute. Lithography is the bottleneck. Source Foundry is the first serious attempt to break it since ASML’s rise. That is the story the market is missing.