A flexible cryogenic cable design from MIT Lincoln Laboratory tackles the unglamorous but critical problem of heat load and wire count inside dilution refrigerators — and could determine which institutions can afford to scale quantum systems.
The hard part of building a useful quantum computer isn't the qubits. It's the wires.
That's the blunt framing from engineers at MIT Lincoln Laboratory who have spent years working on a problem the quantum computing industry rarely talks about in public: dilution refrigerators can only hold so many wires before they warm up and the qubits decohere. As qubit counts climb toward the thousands and beyond, the humble act of threading control and readout lines through a cryogenic system has become one of the field's most stubborn engineering bottlenecks.
On Wednesday, MIT Lincoln Laboratory published details of a prototype flexible, ribbon-like cryogenic cable designed to replace the rigid coaxial cable arrays currently used in most dilution refrigerators. The design — a stripline structure with conductive layers sandwiched between polymer layers — provides electromagnetic shielding, frequency-consistent signal behavior, and minimal heat loss while operating at the millikelvin temperatures required to keep superconducting qubits stable. The group calls the architecture low-frequency, or LF.
"A quantum computer's dilution refrigerator can only hold so many wires before it warms up and the qubits decohere," the research team explains on its group page. The statement underscores a constraint that scales with every new qubit added to a system: each qubit needs at least one control line and typically several readout lines, and each line that penetrates the refrigerator's thermal stages carries heat load into an environment colder than deep space.
The design has been licensed to Maybell Quantum, a Colorado-based company that builds dilution refrigerator systems and quantum computing infrastructure. Maybell is branding the cable as LF CryoTrace and plans to integrate it across all thermal stages of its refrigerators, starting with low-frequency services — thermometry, heaters, and sensors — with qualification testing for additional functions pending.
"As qubit counts scale, coaxial cable arrays become too stiff and bulky, and they add heavy heat loads," said Lasse Nielsen, strategy and operations lead at Maybell Quantum, in a statement accompanying the announcement. "The ribbonized wiring is mechanically robust, reduces handling-related breakages, and turns assembly tasks that took days into tasks done in a few hours."
Maybell founder and CTO Kyle Thompson described the deal as part of the company's broader push to bridge the gap between specialized lab-based quantum environments and commercially viable quantum computing at U.S. manufacturing scale.
The principal investigator on the Lincoln Laboratory side is John Cummings.
Dilution refrigerators used for superconducting quantum computing operate at temperatures between 5 and 10 millikelvins — colder than the void of space. Creating and maintaining that environment requires enormous cooling capacity, and every wire that enters the system represents a heat-leak path. Traditional coaxial cables,bundled in large numbers to service qubit arrays, are stiff, brittle, and physically demanding to install. They also generate friction and handling stress that leads to breakage during assembly and maintenance.
The MIT design uses standard printed-circuit-board fabrication techniques, which the researchers say makes the cables cheaper to produce at scale and easier to route through the tight thermal stages of a dilution refrigerator. The ribbon format also allows higher density: more channels per cross-section, with consistent impedance and crosstalk characteristics that matter for high-fidelity qubit control.
The practical benefit, according to Maybell: what once required days of careful cable installation can now be done in a few hours, using a system designed to survive repeated thermal cycling without degradation.
Quantum computing coverage tends to focus on qubit counts and landmark announcements — a new processor with 1,000 qubits, a coherence time record, a vendor claiming quantum advantage on a narrow task. Less covered is the unglamorous systems engineering that determines whether a laboratory prototype can become a deployable machine.
The wiring bottleneck is one instance of that gap. As superconducting qubit systems have grown from handfuls to hundreds and toward the low thousands, the control and readout infrastructure has become a primary scaling constraint. Vendors including Bluefors, Oxford Instruments, and Leiden Cryogenics have each developed their own approaches to cryogenic wiring and interconnect systems — an indication that the problem is recognized across the industry but has not been solved to everyone's satisfaction.
The MIT cable's novelty, as presented, lies in its compatibility with PCB manufacturing and its ribbon architecture — a departure from the coaxial cable bundles that have dominated. Whether that architecture represents a genuine advance over existing cryostat-vendor approaches, or primarily an economic and handling improvement, will depend on independent benchmarking that has not yet been publicly disclosed.
Maybell is positioning the licensing agreement as a move toward making quantum computing infrastructure more accessible to institutions that cannot afford years of custom engineering per system. The company's stated target is U.S. manufacturing scale-up — a phrase that signals ambition beyond the handful of hyperscalers and national labs that have traditionally been the only organizations able to afford state-of-the-art quantum systems.
For now, the cable is a single component in a much larger stack. The path from a prototype cable to a fully qualified, production-deployed wiring system inside a commercially sold dilution refrigerator is measured in months or years of thermal cycling, signal integrity testing, and integration work. Maybell has not disclosed pricing, availability timelines, or which of its refrigerator models will carry the technology first.
What the announcement does make clear is that the question of who solves the wiring problem — and whether a lab-designed solution can cross over into commercial use — is no longer just an internal engineering concern. It has become a contestable question with real commercial stakes, sitting alongside qubit quality and error correction as one of the variables that will determine which quantum computing platforms actually scale.