Recent breakthroughs in quantum technology, talked through like friends do
We follow what is actually happening in quantum computing, quantum communication and quantum sensing, and we write it up in everyday words. Logical qubits that finally beat the parts they are made from, satellites handing out entangled photons, atomic clocks that can tell which floor of a building they are on, and the long argument over whether anyone has really shown a quantum advantage yet. No maths exam, no breathless press release. Just a good look at the results, with the caveats attached.
Where the field has moved, one result at a time
Each stop on the line is a different corner of quantum technology. We name the groups, the papers and the machines, and we say what was shown and what was not.
- Computing
Logical qubits that beat their parts
Google Quantum AI's below-threshold surface-code result on the Willow chip, the Harvard, MIT and QuEra experiment that ran 48 logical qubits on neutral atoms, Quantinuum's trapped-ion logical qubits, IBM's low-overhead "gross" code and the cat-qubit approach from Alice & Bob and AWS. What each one demonstrated, and what still sits on the to-do list.
Read the error-correction guide - Hardware
Six ways to build a qubit
Superconducting circuits from Google, IBM and Rigetti; trapped ions from Quantinuum and IonQ; neutral atoms from QuEra, Pasqal and Atom Computing; photonics from PsiQuantum and Xanadu; silicon spins from Intel and Diraq; and Microsoft's contested topological programme. We compare cooling, gate speed, connectivity and the scaling story of each.
Compare the platforms - Communication
Quantum networks and the cryptography clock
China's Micius satellite and the Beijing–Shanghai key-distribution backbone, the three-node entanglement network at QuTech in Delft, the EU's EuroQCI programme, and the post-quantum standards NIST published as FIPS 203, 204 and 205. Why the UK's NCSC backs one approach and not the other.
Read the networks guide - Sensing
The quiet breakthroughs that already work
Optical lattice clocks at NIST, JILA, RIKEN, PTB and NPL, the University of Birmingham gravity gradiometer that found a buried tunnel, wearable magnetoencephalography from Nottingham, squeezed light inside LIGO, and the UK flight trials of quantum navigation. Sensing is the part of the field that ships.
Read the sensing guide - Debate
Quantum advantage: claims and pushback
Google's Sycamore sampling experiment and IBM's rebuttal, boson sampling from USTC's Jiuzhang and Xanadu's Borealis, IBM's 2023 "utility" experiment and the tensor-network replies, D-Wave's annealing claim, and the Gidney–Ekerå resource estimates for breaking RSA. How to read the next headline.
Read the advantage guide
Four questions we ask of every quantum announcement
This is our own habit rather than any official standard, but it has saved us from a lot of over-excitement.
Physical or logical?
A count of qubits means very little on its own. A hundred noisy physical qubits and a hundred error-corrected logical qubits are different worlds. We look for which one a release is counting, and whether the logical error rate actually came out below the physical one.
Memory or computation?
Many of the best error-correction results so far store a qubit for longer than a bare qubit would survive. That is a genuine milestone, but it is not running an algorithm. We check whether logical gates were performed, and how many.
Who has tried to match it classically?
Every quantum advantage claim we know of has drawn a classical reply, sometimes within weeks. We wait for the tensor-network and dequantisation crowd to respond before deciding what a result means.
Lab or field?
Especially for sensors and networks, a result in a vibration-isolated basement is not the same as one on a moving vehicle or over deployed fibre. We look for where the thing ran and for how long.