SandboxAQ lands $500M CHIPS Act award to cut foreign chip-material dependence

What's the deal? The US Department of CommerceDealroom has a profile for this one. Try Dealroom → has signed a definitive agreement to award $500M to SandboxAQ, an AI-and-quantum technology startup, under the CHIPS and Science Act. The money will fund AI-driven discovery of alternative materials for domestic semiconductor manufacturing — targeting replacements for PFAS "forever chemicals," new catalysts, rare earth-free magnets, and novel battery chemistries for chip factory backup power.

In an unusual twist, Commerce is taking a minority, non-voting equity stake in SandboxAQ plus future royalty payments if the research yields commercially licensed products. The structure mirrors a recent $2B quantum-computing award and positions Washington more as a venture investor than a traditional grant-maker.

SandboxAQ, backed by Nvidia, Google, Eric Schmidt, and Ray DalioDealroom has a profile for this one. Try Dealroom →, was valued at $5.75B last year. It will use its proprietary "Large Quantitative Models" — trained on physics and chemistry rather than human text — to screen millions of candidate materials in software before anything reaches a lab.

Why now? China controls more than 90% of the rare-earth magnets used in the machines that produce America's most advanced chips. PFAS regulations are tightening worldwide, threatening the supply of chemicals that chipmakers depend on for lithography, heat transfer, and insulation. The US needs domestic alternatives before foreign chokepoints become outright disruptions.

SandboxAQ says its platform can compress discovery timelines from years to weeks. Its catalyst models, built on 13.5 million quantum chemistry calculations developed with Nvidia, claim screening speeds roughly 20,000 times faster than conventional methods.

What could go wrong? This is a research bet, not a product. No new magnet or PFAS-free chemical exists yet — the award funds the search, not a guaranteed result. Translating computational predictions into full-scale manufacturing is a long road with no assurance of commercial viability.

There is also a policy tension: the same administration funding PFAS replacements rolled back some PFAS drinking-water deadlines last year.

The signal: The deal reflects a broader shift in how the US government funds strategic technology. By taking equity stakes and royalty rights rather than issuing grants, Commerce is importing venture capital logic into industrial policy. It also underscores a growing conviction that AI-driven simulation — not just traditional lab work — will determine who controls the next generation of critical materials. If it works, it could reshape how governments everywhere approach supply-chain sovereignty.

Read more: NIST · The Next Web · Quantum Computing Report