Cryogenic computing
Control chips must move into the fridge for quantum to scale, but the layers worth holding are being claimed fast; component suppliers have a short window before margins go.
- Bear
- Open
- $848M by 2035
The callThe restraints win on who keeps the profit: the need for in-fridge control is real, but the foundries, instrument makers and IBM are claiming the design kit and the control box faster than startups are. The market goes from $220M to about $850M by 2035 if quantum machines pass 10,000 qubits on roadmap.
- D1Useful error-corrected machines need millions of qubits, and one coaxial cable per qubit from room temperature cannot scale, so control silicon has to move into the fridge.
- D2Quantum computing is worth $43-71B by 2035 if machines reach that scale, and every machine built needs control silicon.
- R1The defensible layers are already claimed: GlobalFoundries and SEEQC with Cadence hold the design kit, Quantum Machines, Zurich Instruments and the other instrument makers the control box, IBM builds its own.
- R2The near market is small; implied winner revenue is about $44M and SEEQC's SPAC at about $1B is the ceiling comparison.
- R3Silicon carbide electronics working at 10 millikelvin could make the 4-kelvin control architecture obsolete.
- C1A shipping quantum system passes 10,000 physical qubits with control chips inside the fridge, moving control from room-temperature cables to silicon.2031
- 1IBM's roadmap milestones toward 10,000 qubits; that roadmap is the clock.Dec 2029 · open
- ·A cryo component supplier winning a design on GlobalFoundries' Dresden process and keeping margin; a first commercial order from a cryo-CMOS newcomer is the nearest marker.
- ·Silicon carbide electronics confirmed working at 10 millikelvin, which would change the view on the architecture.
- FORMFormFactor
- fair
- GFSGlobalFoundries
- cheap
- GOOGLGoogle
- cheap
- IBMIBM
- full
- INTCIntel
- full
- NVDANVIDIA
- cheap
- KEYSKeysight Technologies
- cheap
- 688027.SSQuantumCTek
- not screened
Private · 12 private companies, 5 of them in the US.
Seeqc · Snowcap · Maybell Quantum · FormFactor (HPD) · CEA-Leti · Bluefors · Rhonexum · Zero Point Cryogenics · VEIR · Emergence Quantum · Quantum Machines · Snowcap Compute
12 more private companies tracked, not shown
$369M raised by 5 companies, $316M of it in the last three years. Wave building: Money is arriving: recent rounds and tier-1 funds are in, but no runaway leader yet.
Backed by a top-tier venture firm: Snowcap Compute. Backed by a strategic investor: Quantum Machines.
- $280MQuantum Machines
- Founded 2018
- $65MMaybell Quantum
- Founded 2021
- $23MSnowcap Compute Seed, Jun 2025 $23M (Playground Global, Cambium Capital, Vsquared Ventures)
- $1MRhonexum
- Founded 2025
- ·Zero Point Cryogenics
- Founded 2017
- 20 Jul 2026The company that makes the fridges every quantum computer needs just sold for less than one good Series C. Enabling layers are not automatically valuable layers.
- 6 Jul 2026A cryocooler founder pins down the near-term cooling market: his machine is being built to cool the warmer, high-power stage (a few watts at four kelvin, proven by late August) rather than the ultra-cold qubit stage the fridge duopoly serves — so the two are complements, not rivals. The honest catch is that almost no customer today needs that much cooling power; the one real pull is a superconducting-compute startup, and an earlier attempt to run ordinary chips cold died because the refrigeration overhead simply was not worth it. Cryogenic computing for classical chips is on no major roadmap for the next five to eight years.
- 6 Jul 2026A cryo-control founder confirms the trap the whole space sits in: the buyers of quantum control want a finished system, not a component, so a chip maker gets squeezed into a supplier margin unless it owns a bigger slice. On the classical side, cooling ordinary chips does not fit where datacentres are heading, which is deliberately warmer water and no chillers, not colder. The one datapoint that could change the superconducting story is a claim, still unverified, that a leading player has finally solved the cryogenic-memory problem that sank every attempt since the 1980s.
8 earlier
- 24 Jun 2026A North American maker of cryogenic refrigerators reframes where the near-term money in cryo cooling actually is: not fault-tolerant quantum computing (still too small and slow to supply at scale), but quantum SENSING — multi-million-dollar government projects, including a late-stage Arctic-sensing contract whose tender was written around the company. The standout commercial signal is a non-dilution "phase-separation" fridge, now their fastest-adopting product, claiming ~30% lower cost and ~80% less helium than a dilution fridge — a cost attack on the cooling layer that is already shipping rather than slideware.
- 21 Jun 2026A wildcard in cryogenic control: a Hong Kong group made a standard silicon-carbide transistor behave like a spiking neuron at 10 millikelvin — the qubit's own temperature — without the heat that forces everyone else to keep control electronics one cold stage away and wire down. If it scales it would route around the dominant approach, though it is still a single-device lab result.
- 18 Jun 2026Mapping the cryogenic-computing field across its three layers — the cooling hardware, the cold control electronics, and the cold compute itself — one cooling player stands apart: instead of building a better helium dilution fridge, it cools electronically with solid-state junctions, claiming a roughly 20x lighter, far lower-power box with no scarce helium-3. It is the most disruptive and least mature bet on the cooling layer, and like its peers it is still grant-funded and pre-company rather than venture-raising.
- 16 Jun 2026US government money has turned cryogenic chip manufacturing for quantum into a funded national priority: a $2bn programme backing two quantum foundries and seven quantum-computing companies, with the merchant foundries now explicitly committing to cryogenic design kits for the control and read-out chips. That hardens the why-now for the near-term cryo-control market, but it also means the most defensible layer (the cryo process and design kit) is being claimed by the giants, leaving smaller players to compete on differentiated components that ride on top.
- 16 Jun 2026A North American maker of dilution refrigerators — one of only about six firms worldwide that commercially build the fridges quantum computers need — is raising a pre-Series A to challenge the European near-duopoly that dominates cryogenic cooling. It rounds out the cooling part of the theme: alongside the pre-commercial cryocooler builders, there is now a shipping merchant challenger, though the open question is unchanged — whether the cooling overhead ever pays for itself, and whether a new entrant can hold margin against entrenched incumbents.
- 15 Jun 2026A cryo-cooling hardware founder confirms the desk-research read: across competing cryo-compute architectures, nothing has reached commercial fruition yet, but every prospective customer names the same cooling bottleneck. The cooling-layer builders are still grant-funded and deliberately not venture-raising, waiting to see which chip approach wins.
- 15 Jun 2026A European cryo-CMOS component startup reports quantum hardware makers shifting control electronics from in-house to external suppliers, and that the realistic route to market for a small component maker is integrating into an established control-system vendor's stack rather than selling direct. The gating technical bottleneck is the absence of reliable cryogenic device models from the foundries, which a wave of EU quantum-design funding is now aiming squarely at.
- 10 Jun 2026Cryo-CMOS control for quantum is accelerating — a new merchant quantum-foundry line, fault-tolerance roadmaps to 2029, and the first commercial cryo-CMOS orders shipping — but the control box is already owned by incumbents, so the only venture-shaped opening is the hard cryo component (switches, multiplexers, low-noise amplifiers) taped out on a commercial foundry process.