A study published on August 19 in *Physical Review Letters* by J.J. Thio and David Arvidsson-Shukur's team at the University of Cambridge's Cavendish Laboratory draws a sharper boundary around a famous slogan in quantum computing: "magic is all you need." Magic states are considered the fuel of quantum computation—without them, a quantum computer is no more powerful than a classical one. The Cambridge result shows that magic is a necessary condition, but not a sufficient one—there is a hard edge between "useful magic" and "useless magic."
Key insight: Kirkwood-Dirac negativity
The Kirkwood-Dirac (KD) distribution was born in Cambridge in 1945: physicist Paul Dirac, at St John's College, extended a similar distribution proposed by John Kirkwood (MIT) in 1935 into a complete mathematical framework. The framework contains a seemingly paradoxical concept—negative probabilities. Dirac insisted that negative solutions in equations should be taken seriously, a stance that eventually led to the prediction of antimatter.
The Cambridge team dug up this 80-year-old mathematical tool and gave it a new purpose:
- When the KD distribution over a quantum state remains positive or zero throughout the entire computation, a classical computer can simulate it easily.
- When the KD distribution takes negative values, the difficulty of classical simulation climbs exponentially—often signaling that genuine quantum advantage may be possible.
On-the-spot validation: a 'quantum task' completed on a laptop
To ground the theory, Thio, working with fellow PhD student Rishi Goel, wrote a classical simulation program on an ordinary laptop. The program successfully completed computational tasks previously widely believed to require a quantum computer.
In other words, the boundary of classical computation is wider than previously assumed. Arvidsson-Shukur put it bluntly: "Without this kind of foundational research, we would never truly be able to judge whether a quantum device has actually achieved a breakthrough that genuinely exceeds the reach of classical computation."
Dr. Nicole Yunger Halpern of the Joint Quantum Institute and Joint Center for Quantum Information and Computer Science at the University of Maryland commented that some researchers had suspected for years that KD negativity might help identify genuinely useful magic states, and she is excited that this team has finally given a positive answer.
Why this matters
Governments and private companies worldwide are pouring tens of billions into quantum computing. But when quantum computers will actually beat classical machines on practically valuable tasks remains highly uncertain—partly because the definition of "quantum advantage" itself is vague.
The Cambridge team's contribution is not another criterion for claiming quantum advantage, but a negative test for such claims: if you cannot produce a proof that KD negativity exists, your "quantum advantage" claim lacks theoretical support. This imposes a new engineering checkpoint on every paper and company claiming quantum advantage.
It also matters deeply for magic-state preparation—one of the most critical engineering challenges in building large-scale quantum computers. Identifying "useful" magic states more precisely directly determines whether such resources can be manufactured and applied. It is a set of navigation coordinates for this engineering problem.
A judgment: redrawing the quantum-advantage discourse
Over the past decade, "quantum supremacy / quantum advantage" has become an industry buzzword. Whoever runs a circuit on 50, 100, or 200 qubits that "classical computers can't simulate" makes headlines. But many such "advantages" are fragile—progress in classical algorithms or the discovery of mathematical structure can suddenly erase them (see IBM's tensor-network improvement that undermined Google Sycamore's quantum-advantage claim).
The Cambridge team offers not a list of "which algorithms can beat classical," but a reverse criterion: prove that the KD distribution goes negative first—only then may you discuss quantum advantage. This redraws the discourse: "magic" is no longer a claim, but a proof.
Who will be affected
1. All quantum computing companies—especially those marketing "quantum advantage"—will need to supply "KD negativity evidence." 2. Classical algorithm researchers—tasks previously assumed to require quantum hardware can now follow the "KD non-negative → classically simulable" path. 3. Government technology policy—multi-billion-dollar investments will be re-examined through the sieve of KD negativity.
The word "magic," from now on, must pass the KD negativity test before it appears in a paper.
References: Physical Review Letters paper (Aug 19); QuantumWire (Aug 19); NetEase Tech/163.com translation (Aug 23); The Qubit Report Weekly Round-Up (Aug 22), "Verification" column; J.J. Thio (St John's College PhD); David Arvidsson-Shukur (Hitachi Laboratory, Cavendish); Prof. Crispin Barnes's group; Nicole Yunger Halpern (JQI, Maryland).