2025 · Clarke, Devoret & Martinis: The Quantum Computing Nobel

On 7 October 2025, the Royal Swedish Academy of Sciences announced a Nobel Prize this course had been waiting for. John Clarke, Michel Devoret and John Martinis won the Physics prize "for the discovery of macroscopic quantum mechanical tunnelling and energy quantisation in an electric circuit." In plain words: they proved that billions of electrons in a hand-held electric circuit can behave as a single quantum particle — and every superconducting qubit on Earth descends from that discovery.

The story starts in 1984, in John Clarke's laboratory at UC Berkeley. The team was small: Clarke himself, the professor and SQUID pioneer; Michel Devoret, a young postdoc visiting from Saclay in France; and John Martinis, Clarke's own PhD student. The orthodoxy of the day said quantum mechanics was for atoms and photons — a circuit on your workbench was classical, full stop. They set out to prove the orthodoxy wrong.

Their device was a Josephson junction — two superconductors separated by an insulating layer so thin that electron pairs can tunnel straight through it — wired into a circuit and cooled to about 0.01 kelvin, colder than deep space. They demonstrated two things. First, the circuit could escape from its zero-voltage state by quantum tunnelling — leaking through an energy barrier that no classical physics could cross. Second, the circuit's energy levels were quantized: discrete steps, like an atom's, not a smooth continuum. Their masterstroke was a 1984 calibration trick called resonant activation — worked out with Daniel Esteve, a second visitor from Saclay in France: they used microwaves to tickle the circuit between its energy levels, proving the levels were real — with data so clean it matched Caldeira–Leggett tunnelling theory with no fitted parameters at all. To kill thermal noise, they built filters that damped microwave noise between room temperature and the sample by twenty orders of magnitude.

The paper trail is unusually clean. 1984 — Devoret, Martinis, Esteve & Clarke, Physical Review Letters 53, 1260: the resonant-activation calibration. 1985 — Martinis, Devoret & Clarke, PRL 55, 1543: the first quantized energy levels of a macroscopic variable, published on 7 October 1985. 1985 — Devoret, Martinis & Clarke, PRL 55, 1908: the tunnelling rates matched theory with every parameter measured in place. 1988 — Clarke and colleagues' synthesis in Science 239, 992. All of it done in Clarke's group at UC Berkeley and Lawrence Berkeley Lab — Martinis as the PhD student, Devoret as the postdoc on leave from Saclay. The Nobel was announced on 7 October 2025: exactly forty years to the day after the key paper. Nobody plans that; the committee doesn't work that way. It is just how the arithmetic of history landed.

Why this is the quantum computing Nobel: the trio had built the first engineerable artificial atom — a circuit you can design on paper that behaves like an atom. The direct line runs from Berkeley 1985 to the Cooper-pair box (1997), the first superconducting qubit (1999, coherence about 10 nanoseconds — Nakamura, Pashkin & Tsai at NEC), flux qubits holding superpositions of clockwise and anticlockwise persistent currents (2000), the phase qubit and quantronium (2002), circuit QED (2004), the transmon (2007, Devoret a co-author — the chip inside Google's and IBM's processors today), and fluxonium pushing coherence toward milliseconds. The human line is just as direct: Martinis joined Google in 2014 and led the 2019 quantum supremacy experiment — 53 transmon qubits doing in about 200 seconds what was estimated at 10,000 years classically — then left in April 2020 and co-founded the superconducting-quantum startup Qolab as CTO in 2022. Devoret is now Chief Scientist for Quantum Hardware at Google Quantum AI, and Google's Willow chip (2024/25) has since shown quantum error correction below threshold. Along the way the trio shared the Fritz London Memorial Prize (2014) — the classic pre-Nobel signal.

The prize itself: announced 7 October 2025 in Stockholm — the discovery presented by Nobel Committee chair Olle Eriksson — and presented 10 December 2025 at the Stockholm Concert Hall by King Carl XVI Gustaf, with the physics presentation speech by Professor Göran Johansson and the banquet speech given by Michel Devoret himself. The money: 11 million SEK, one third each — about 3.67 million SEK apiece. The laureates: John Clarke (born 1942, Cambridge, UK; PhD 1968; the SQUID pioneer; emeritus at Berkeley and Lawrence Berkeley Lab) — 83 at the announcement; Michel Devoret (born 1953, Paris; PhD 1982, Paris-Saclay; Yale, UC Santa Barbara, Google Quantum AI) — 72; John Martinis (born 1958; PhD 1987 at Berkeley under Clarke; UC Santa Barbara) — 67, also a John Stewart Bell Prize winner (2021) and one of Nature's 10 (2019). Clarke's reaction: "To put it mildly, it was the surprise of my life… it never occurred to me in any way that this might be the basis of a Nobel Prize." The prize took Berkeley Lab's count of associated Nobel laureates to 17. And the timing: the centenary of quantum mechanics (Heisenberg, 1925) and the UN's International Year of Quantum Science and Technology. One honest footnote — the Nobel committee never said the quantum year influenced the choice; that timing link is commentators' inference, not a committee statement.

Common myth: "They won for inventing the qubit." They didn't — the first superconducting qubit came in 1999, fourteen years after their experiment. They discovered the quantum behavior of electric circuits that made qubits possible. The Nobel rewards the discovery, not the engineering that followed.