Situational Awareness Bets $400M on AI Chip Startup Source Foundry

Leopold Aschenbrenner’s AI-focused hedge fund Situational Awareness has made one of its biggest private technology bets yet, investing another $400 million in semiconductor startup Source Foundry even as the fund recovers from a bruising selloff in AI-related public stocks.

The latest investment takes Situational Awareness’s total commitment to Source Foundry to about $500 million, according to The Wall Street Journal. The startup, founded in 2025 by Stanford researchers Abdulmalik Obaid and Joe Burg, has reportedly raised several hundred million dollars and was valued at approximately $5 billion in a recent financing. Sequoia Capital is also among its previous backers.

What makes the investment particularly significant is where Source Foundry is positioning itself. Rather than designing another GPU or AI accelerator, the company is reportedly developing technology intended to improve the machines and manufacturing processes used to produce advanced semiconductors. That puts it much closer to the foundation of the AI hardware supply chain.

Source Foundry Is Going After a Critical Chipmaking Bottleneck

Source Foundry remains largely in stealth, but reporting around the company points to semiconductor lithography as a central part of its ambitions. Lithography systems project extraordinarily detailed circuit patterns onto silicon wafers, making them essential to producing increasingly advanced processors.

The market is particularly difficult to enter because Dutch semiconductor-equipment giant ASML has spent decades developing the technology and supply chain behind advanced lithography.

ASML's scale shows what Source Foundry would eventually be competing against. The company generated €32.7 billion in net sales in 2025, invested €4.7 billion in research and development, and shipped 48 extreme-ultraviolet, or EUV, lithography systems during the year. Its supplier network stretches across roughly 5,100 companies worldwide.

AI infrastructure demand is pushing those numbers higher. ASML reported €9.3 billion in net sales during the second quarter of 2026 and raised its full-year sales forecast to between €43 billion and €45 billion, with a projected gross margin of 54% to 56%.

That makes Source Foundry's reported $5 billion valuation aggressive for a company founded only last year, but it also illustrates how investors increasingly view semiconductor manufacturing equipment as a strategic AI asset rather than a niche industrial business.

The $400 Million Check Comes After a Painful Hedge Fund Reversal

The timing makes the investment even more unusual.

Situational Awareness recently suffered steep losses as AI infrastructure stocks reversed sharply. The fund was hit by margin calls and ultimately sold a large portion of its public-equity portfolio to Citadel. Reuters reported that some of those positions were transferred at discounts exceeding 10%.

TechCrunch reported that Situational Awareness's assets under management fell from roughly $20 billion to around $10 billion following the turmoil. The fund nevertheless retained private holdings including its position in Anthropic.

Other reporting has put the fund's July decline at about 67%, highlighting how quickly its highly leveraged exposure to AI infrastructure turned against it.

Against that backdrop, committing another $400 million to an early-stage semiconductor company looks less like a retreat from Aschenbrenner's AI thesis and more like a change in where he wants to express it.

Instead of concentrating entirely on publicly traded companies whose valuations can swing rapidly with market sentiment, Source Foundry represents a longer-term infrastructure bet on the machinery required to keep AI computing capacity expanding.

AI Investment Is Moving Below the GPU Layer

For much of the generative AI boom, Nvidia GPUs and cloud data centers dominated the investment conversation. Increasingly, however, capital is moving deeper into the semiconductor supply chain.

Lithography, advanced packaging, memory, power infrastructure and semiconductor fabrication capacity have all become strategically important as companies attempt to build larger AI clusters.

ASML itself has said that semiconductor demand continues to strengthen because of AI infrastructure investment, with customers accelerating capacity expansion plans for 2026 and beyond. In the first quarter, the company reported €8.8 billion in sales and initially raised its 2026 revenue outlook before increasing that forecast again after the second quarter.

That demand helps explain why investors may be willing to finance unusually ambitious semiconductor startups despite the enormous technical barriers.

Source Foundry is still far from proving it can challenge established equipment suppliers. Building semiconductor manufacturing tools requires extreme precision, massive R&D spending, customer qualification and a complex network of specialist suppliers. ASML's existing ecosystem alone includes thousands of suppliers and decades of accumulated engineering knowledge.

Still, Situational Awareness's $500 million total commitment gives Source Foundry considerably more capital than most young hardware companies receive at such an early stage.

A High-Risk Bet on the Next Layer of the AI Boom

The Source Foundry investment ultimately reinforces Aschenbrenner's broader conviction that the AI buildout will require enormous investment not only in models and software, but also in the physical infrastructure beneath them.

The difference is that this latest wager targets the manufacturing layer itself.

If Source Foundry can materially reduce the cost, improve the speed or expand the availability of advanced chip-production technology, its opportunity could extend well beyond one generation of AI accelerators. But challenging companies that have spent decades mastering semiconductor equipment will require far more than capital.

For Situational Awareness, the deal is therefore both a technology bet and a test of investment strategy. Just days after one of the most dramatic reversals in the fund's short history, Aschenbrenner is again placing hundreds of millions of dollars behind the assumption that AI's biggest bottlenecks will increasingly be found in hardware.

And this time, the wager goes several layers deeper than the chips themselves.