Researchers at Chalmers University of Technology have built a quantum refrigerator.

It uses noise to move heat on demand inside superconducting quantum circuits.

The work shows that random microwave fluctuations can be engineered to pump heat, regulate it.

They can even amplify thermal flow at scales that conventional cooling cannot touch.

This matters because superconducting quantum computers live on the edge of absolute zero. At roughly minus 273 degrees Celsius, resistance disappears and qubits can hold quantum states.

The catch is brutal. Heat and stray signals kill those states fast. As chips scale from tens of qubits to thousands, heat spreads internally and cryogenic hardware struggles to keep up.

Chalmers’ device attacks that internal heat problem directly. Instead of relying on bulky mechanical refrigeration, the team built a superconducting artificial molecule connected to hot and cold microwave reservoirs.

A third channel injects controlled noise, random microwave signals in a narrow band. That noise acts as the switch. Turn it on and heat flows. Shape it and heat moves in specific directions. Turn it off and the system locks down.

The disruption behind the news: Temperature becomes programmable.

Right now, quantum hardware depends on a stack of dilution refrigerators that cool the whole chip uniformly.

That approach does not scale well.

As processors grow, hot spots emerge during computation, and there is no fine control to drain heat locally.

Noise has been treated as poison to be eliminated at all costs.

Chalmers just showed the opposite. Noise can be an active component. Inject the right spectrum and you get directed heat flow through a circuit that behaves like a molecule. That changes the architecture playbook.

The system can switch between cooling and heat amplification modes, and requires no moving parts. That combination is new. It means future quantum chips could include on-chip thermal routers that stabilize qubits during operation instead of shutting everything down when temperatures drift.

This is the first credible path toward thermal management that scales with qubit count instead of against it. If you can control heat the same way you control signals, you unlock denser layouts, longer coherence times, and higher duty cycles.

This is a prerequisite for useful quantum machines.

What to watch next

Over the next 6 to 24 months, watch for three things.

First, integration.

The big test is whether noise-driven refrigerators can be embedded directly into multi-qubit chips without introducing new failure modes. If they can, hardware roadmaps change fast.

Second, costs.

Mechanical cryogenics are expensive and power-hungry. On-chip thermal control shifts part of that burden into fabrication and microwave electronics. That favors companies that already own advanced superconducting process lines.

Third, competition spillover.

Techniques like this will not stay confined to quantum computing. Ultra-sensitive detectors, space instrumentation, and low-noise communications all fight the same heat problem at small scales.

Anyone building large superconducting quantum systems without internal heat control is taking a gamble. They’re betting physics will stay forgiving. It never does. Noise just switched sides, thanks to the quantum refrigerator. Teams that ignore this will fall behind fast.