Cryogenics & dilution refrigerators

Every commercial superconducting quantum processor — IBM's, Google's, Rigetti's, IQM's — does its work at the bottom of a dilution refrigerator, below 20 millikelvin, colder than deep space. The machines that get there come overwhelmingly from two franchises: Bluefors of Helsinki, and the Oxford Instruments NanoScience line (Proteox, Triton) — which just changed hands. Oxford sold NanoScience for £60m to San Diego's privately held Quantum Design, a deal completed January 2, 2026. Read that twice: the only direct listed exposure to this chokepoint exited public markets six months ago — and even that was one ~£59m division inside a diversified ~£500m instruments group, never a pure-play.

The moat is tacit process knowledge, not patents. A dilution unit is silver-sinter heat exchangers, helium-3/helium-4 gas handling, and vibration engineering refined over decades — and the leader keeps deepening it through vertical integration. Bluefors acquired Cryomech, one of only two scaled makers of the low-vibration pulse-tube cryocoolers that precool every fridge to 4 kelvin; the other is Sumitomo Heavy Industries. The market leader in fridges now owns one of the two scaled suppliers of its competitors' most critical component. Above that sits an input choke nobody engineers around: helium-3, harvested from decaying tritium stockpiles, with terrestrial supply estimated around 22,000–30,000 liters a year against demand of 40,000–60,000 — see also War on the Rocks calling dilution refrigerators the single most critical quantum bottleneck. The franchise position is visible at the top of the market: IBM Quantum System Two runs on Bluefors' KIDE platform, three dilution units cooling a 500 kg payload.

Be honest about scale. NanoScience — the global number two — generated roughly £59m of revenue and £1m of adjusted operating profit in FY25, and the whole franchise cleared for less than the price of a single EUV tool. The chokepoint is real and the market is small; you are underwriting structure ahead of volume, not current earnings.

What breaks it? Three vectors. First, capitalized challengers: Maybell Quantum has raised $65m from Addition, Cerberus and In-Q-Tel for radically denser, smaller-footprint fridges. Second, adjacent roll-ups: FormFactor quietly became the largest US-based dilution-refrigerator supplier by buying HPD and the Janis ULT dilution-refrigerator line — today the only way to own this choke on NASDAQ. Third, demand-side bypass: trapped-ion, neutral-atom and photonic modalities don't need millikelvin, and IBM's in-house Goldeneye cryostat is the standing reminder that a determined buyer can self-supply.

The repricing catalyst is procurement, not physics. DARPA's Quantum Benchmarking Initiative moved 11 companies into Stage B in November 2025, with Stage C down-selects expected late 2026 — every superconducting finalist implies multi-year fridge and cryocooler orders. In parallel, the National Quantum Initiative Reauthorization Act cleared House Science on April 29, 2026 and is headed to the floor: five more years of authorized lab funding is the demand base for this entire supply chain. Watch FORM's quarterly Systems commentary as the only public tape on fridge demand — and watch whether Quantum Design, now holding the number-two franchise, ever gives public investors a way back in.