Nobel Prize in Physics 2025: The Quantum Circuit Behind the Computer Revolution

Science & A.I. · Quantum physics

In 1985, in a basement laboratory at Berkeley, three physicists watched an electrical circuit the size of a fingertip do something only individual atoms were supposed to do. Forty years later, that experiment won the Nobel Prize in Physics.

On 7 October 2025 the Royal Swedish Academy of Sciences awarded the prize to John Clarke, Michel H. Devoret and John M. Martinis. Their citation reads: “for the discovery of macroscopic quantum mechanical tunnelling and energy quantisation in an electric circuit”.

Stripped of jargon, they showed that the strange rules of quantum mechanics — usually confined to single particles — can govern an object large enough to hold in your hand. That demonstration is the foundation on which today’s quantum computers are built. This article explains what they found, why it took four decades to be honoured, and what it unlocked.

It is a story with an unusually clean shape. A deep question in fundamental physics, answered with a careful experiment in the mid-1980s, sat quietly for a generation — and then turned out to be the seed of one of the most closely watched technologies of our time.

What the 2025 Nobel was awarded for

The prize recognises two related discoveries in a single device: macroscopic quantum tunnelling, and the quantisation of energy in an electric circuit. Both had been seen before in individual particles. The laureates showed them in a circuit built from superconductors — a system vastly larger than an atom, yet still obeying quantum rules.

At the heart of the experiment sat a Josephson junction: two superconductors separated by a wafer-thin barrier of non-conducting material. In a superconductor, electrons pair up and flow without resistance, and a Josephson junction lets that collective quantum state be probed and controlled electrically. It is the component that made everything else possible.

The importance is conceptual as much as technical. Quantum mechanics was long assumed to fade away at everyday scales, its effects averaged into ordinary classical behaviour. The Berkeley work proved that a carefully engineered object, large enough to be handled, could be coaxed into behaving as a single quantum entity — the essential precondition for a quantum bit.

Quantum on a human scale: tunnelling and quantised energy

Quantum tunnelling is the phenomenon by which a particle passes through an energy barrier it classically should not be able to cross. Picture a ball that lacks the energy to roll over a hill, yet somehow appears on the far side. For single particles this is routine physics; the surprise was seeing it in a whole circuit.

Energy quantisation is the second quantum signature. A classical system can hold any amount of energy along a smooth continuum. A quantum system can only occupy specific, discrete levels, like rungs on a ladder with nothing in between. The laureates demonstrated that their macroscopic circuit had exactly these discrete energy states.

Together, these two results were decisive. A device that tunnels and that has distinct energy levels can be prepared in a chosen state, nudged between levels, and read out — the full repertoire a computer needs from its most basic element. The circuit was, in effect, an artificial atom that engineers could design rather than merely find in nature.

Why superconductors, and why so cold?

The choice of superconductors was not incidental; it was the whole trick. In an ordinary wire, countless electrons jostle and scatter, and their individual quantum behaviour is washed out into the familiar classical flow of current. To see quantum effects in a macroscopic object, that disorder has to be tamed.

A superconductor tames it. Below a critical temperature, electrons bind into pairs and settle into a single, shared quantum state that flows without resistance. The many particles begin to act as one, and that collective state can behave quantum-mechanically on a scale far larger than a single atom — which is precisely what the laureates exploited.

This is also why quantum computers must be kept so cold. The superconducting qubits that descend from the Berkeley circuits operate at temperatures near absolute zero, colder than the depths of space, inside elaborate refrigerators. Warmth is noise, and noise destroys the fragile quantum state before it can be used.

The 1984–85 experiments that started it all

The experiments were carried out at the University of California, Berkeley, in 1984 and 1985. John Clarke led the laboratory; Michel Devoret and John Martinis were the younger researchers who built and measured the circuits. The work demanded extreme conditions: temperatures near absolute zero, and shielding against the faintest electrical noise that could destroy the delicate quantum state.

What made the results credible was their quantitative precision. The team did not merely observe suggestive behaviour; they measured it against the predictions of quantum theory and found agreement. That rigour turned an intriguing possibility into an established fact that others could build upon.

The measurements themselves were ingenious. To reveal tunnelling, the researchers biased the junction close to the point where its quantum state should escape, then timed how often it did so, comparing the statistics against theory. To reveal the discrete energy levels, they used microwaves to nudge the circuit between states, watching it respond only at specific, telltale frequencies.

At the time, the practical payoff was not obvious. The experiments answered a deep question in fundamental physics: how large can a quantum object be? The answer — larger than anyone had shown — would only reveal its full significance decades later, when the technology to exploit it matured.

The question the experiment was built to answer

The Berkeley experiments did not happen in a vacuum. They were designed to test a sharp theoretical question that had been posed in the early 1980s: does quantum mechanics still apply when a system contains a vast number of particles, or does some new physics take over at large scales?

Theorists had worked out how a large object coupled to its noisy environment should behave if quantum mechanics held all the way up. The prediction was specific and testable: a macroscopic circuit should still tunnel, and should still show discrete energy levels, provided the experiment was clean and cold enough to keep the environment from washing the effect away.

That is what makes the result so satisfying. It was not a lucky accident but a confirmation of a bold claim about the reach of quantum theory. The universe did not switch to a different rulebook for large objects; it kept the quantum one, and the laureates proved it in the laboratory.

Who the laureates are

John Clarke, born in Cambridge, England, spent his career at Berkeley and is a leading authority on superconducting devices and the exquisitely sensitive magnetic detectors known as SQUIDs. He was the senior figure whose laboratory made the discoveries possible.

Michel Devoret, born in France, went on to become a central figure in quantum-circuit research, later working at Yale and the University of California, Santa Barbara. His work helped turn the early circuits into the controllable building blocks used in modern quantum processors.

John Martinis, an American physicist, became one of the field’s most prominent experimentalists and led the Google team that reported a landmark quantum-computing milestone in 2019. The arc of his career — from a Berkeley graduate student to the frontier of industrial quantum computing — mirrors the journey of the field itself.

From laboratory curiosity to the qubit

The straight line from the 1985 experiments to today runs through the superconducting qubit. A qubit is the quantum equivalent of a classical bit, but instead of being fixed at zero or one it can exist in a combination of both states at once, a property called superposition. The artificial atom the laureates built is exactly the kind of system that can hold such a state.

Most of the best-known quantum computers — including those built by IBM and Google — use superconducting qubits based on Josephson junctions, cooled to temperatures colder than deep space. Every one of them is a direct descendant of the circuits measured at Berkeley four decades ago.

The refinement from those first circuits to a modern qubit took decades of further work, much of it by the laureates and their students. Designs were made less sensitive to noise, easier to control and more reproducible, so that many could be built to behave alike. But the physical principle never changed: a Josephson junction, held cold and quiet, coaxed into acting as a single quantum object.

This is why the 2025 award was described as honouring both a past achievement and a present revolution. The committee recognised foundational physics, but the timing acknowledged that the technology it enabled had at last become real, moving from the laboratory bench toward genuine computational use.

The state of quantum computing today

The superconducting qubit is not the only route to a quantum computer, but it is the most industrially developed. IBM and Google have built their machines around it, steadily increasing qubit counts and improving the quality of each one. Both have published multi-year roadmaps aimed at large, error-corrected systems.

In 2019 a Google team led by John Martinis reported that its superconducting processor had performed a specific, narrowly defined task faster than the best classical supercomputers of the day could match — an early, contested, but symbolically important demonstration that quantum hardware could do something classical hardware could not. It was a milestone on the road the Nobel work had opened.

Other approaches compete seriously. Trapped-ion machines encode qubits in individual charged atoms; photonic systems use particles of light; neutral-atom arrays hold atoms in place with lasers. Each has strengths and weaknesses, and it is not yet clear which will scale best. What they share is the same underlying goal the laureates made conceivable.

It is worth being honest about the state of play. Today’s machines are noisy and limited, useful for research and narrow demonstrations rather than everyday problems. The gap between a laboratory processor and a computer that reliably outperforms classical machines on commercially valuable tasks remains wide, and closing it is the work of the coming years.

What has changed is the seriousness of the effort. Governments and technology companies now invest heavily in quantum research, treating it as strategically important rather than speculative. That shift, more than any single processor, is what the 2025 prize implicitly acknowledged: a field that was once niche has become a global race.

Why the prize came now

Nobel Prizes often arrive decades after the work they honour, once the significance is beyond dispute. In this case the wait was instructive. For years, macroscopic quantum circuits were a specialist curiosity; only as quantum computing grew into a global research effort did the foundational nature of the Berkeley experiments become impossible to overlook.

By the mid-2020s, quantum processors were being used to tackle specific problems in ways that classical machines struggled to match, and companies and governments were investing heavily in the technology. Against that backdrop, recognising the physicists who proved quantum behaviour could be engineered at scale was both overdue and pointed.

The lesson is a familiar one in science: fundamental research pursued for its own sake, with no application in view, can decades later become the bedrock of an entire industry. The 2025 physics prize is a case study in that long, unpredictable payoff.

What a quantum computer could actually do

The excitement around quantum computing rests on a handful of problems where it could offer a genuine advantage. The most famous is factoring very large numbers. A sufficiently powerful quantum computer running an algorithm devised by the mathematician Peter Shor could break the encryption that secures much of today’s internet traffic.

That prospect, though still distant, has already prompted a worldwide effort to develop new, quantum-resistant forms of cryptography before the threat becomes real. It is a rare case of a technology reshaping security policy years before it exists in usable form.

The more constructive promise lies in simulating nature itself. Molecules and materials are quantum systems, and a quantum computer is naturally suited to modelling them. That could accelerate the design of new drugs, catalysts, batteries and materials in ways classical machines find prohibitively expensive.

Optimisation is a third area, from logistics to finance, where quantum methods might one day find better solutions than classical ones. In every case the honest caveat is the same: the potential is real, the timeline is uncertain, and the hardware is not yet good enough. The foundations, however, are no longer in doubt.

What comes next for quantum technology

The central challenge now is not proving that quantum circuits work but making enough of them work together reliably. Qubits are fragile: the faintest disturbance can collapse their delicate states, a problem known as decoherence. Building useful machines means correcting these errors faster than they accumulate.

Quantum error correction spreads the information of one reliable logical qubit across many physical ones, so that errors can be detected and fixed without destroying the computation. Scaling from today’s processors to the large, fault-tolerant machines that could break cryptography or design new materials remains the field’s defining engineering problem.

Whatever route succeeds, its foundations were laid in that Berkeley basement. The laureates answered a question about the nature of reality — how far into our everyday world the quantum rules reach — and in doing so handed engineers the raw material for a new kind of computer.

There is a quiet lesson in the four decades between the experiment and the prize. None of the three physicists set out in 1984 to build a computer. They were chasing a fundamental question about how the world works, and the answer turned out to be worth an industry.

It is a pattern that recurs throughout the history of science, from electromagnetism to the laser: curiosity-driven research, undertaken with no product in mind, laying foundations that only later prove indispensable. The 2025 Nobel Prize in Physics is both a celebration of a specific discovery and a reminder of why such patient, open-ended inquiry deserves support.

Beyond computing: quantum sensing

Quantum circuits are not only about computation. The same extreme sensitivity that makes a qubit fragile also makes it an exquisite sensor. John Clarke built much of his career on superconducting quantum interference devices, or SQUIDs, which can detect magnetic fields of almost unimaginable faintness.

Such devices are already used to measure the tiny magnetic signals produced by the human brain and heart, to search for signatures of exotic physics, and to probe materials in the laboratory. Quantum sensing may reach practical, everyday use sooner than quantum computing does, because it demands fewer perfectly coordinated qubits.

This breadth is part of why the 2025 prize resonated. The physics it honoured did not open a single door but a whole corridor of them — computing, sensing, secure communication and precision measurement all trace back to the ability to control quantum states in engineered devices.

Frequently asked questions

Sources

Primary peer-reviewed research (the laureates’ own work):

• Devoret, M. H., Martinis, J. M., & Clarke, J. (1985). Measurements of macroscopic quantum tunneling out of the zero-voltage state of a current-biased Josephson junction. Physical Review Letters, 55, 1908. doi.org/10.1103/PhysRevLett.55.1908
• Martinis, J. M., Devoret, M. H., & Clarke, J. (1985). Energy-level quantization in the zero-voltage state of a current-biased Josephson junction. Physical Review Letters, 55, 1543. doi.org/10.1103/PhysRevLett.55.1543
• Clarke, J., Cleland, A. N., Devoret, M. H., Esteve, D., & Martinis, J. M. (1988). Quantum mechanics of a macroscopic variable: the phase difference of a Josephson junction. Science, 239, 992. doi.org/10.1126/science.239.4843.992
• Arute, F., et al. (2019). Quantum supremacy using a programmable superconducting processor. Nature, 574, 505. doi.org/10.1038/s41586-019-1666-5

Nobel Prize documentation:

• The Nobel Prize in Physics 2025 — official press release (Royal Swedish Academy of Sciences)
• NobelPrize.org — Physics 2025 summary and citation
• NobelPrize.org — Popular science background: quantum properties on a human scale

Scientific background and journalism:

• Royal Swedish Academy of Sciences — Scientific background to the 2025 physics prize (PDF)
• NobelPrize.org — Laureate facts: Clarke, Devoret and Martinis