ScienceExplain

How a Superconducting Qubit Actually Works

Advanced

1. Quick Summary

A superconducting qubit is an electrical circuit made from superconducting material, in which current flows with no resistance. Its quantum behaviour comes not from exotic material physics alone but from deliberately introducing a nonlinear element called a Josephson junction.

How a Superconducting Qubit Actually Works
A network: connected nodes passing things along.

The junction is a thin insulating barrier between two superconductors. Electrons cross it without resistance, and its response is nonlinear, which turns what would otherwise be a plain oscillator into something with distinct, addressable energy levels.

Engineers pick two of those levels to serve as zero and one. Microwaves drive transitions between them, and readout is performed by coupling the circuit to a resonator whose response shifts depending on the qubit state.

The whole assembly sits in a dilution refrigerator at millikelvin temperatures. That is not mainly about superconductivity — it is about keeping thermal energy far below the energy spacing between the two levels, so the qubit stays where you put it.

2. What It Means

A plain inductor-capacitor circuit has evenly spaced energy levels. That uniformity makes it useless as a qubit, because a drive tuned to the zero-to-one transition also excites one-to-two and beyond. The Josephson junction makes the spacing uneven, so a carefully tuned microwave pulse addresses only one transition.

This is why the field talks about an ‘artificial atom’. The circuit has discrete levels like an atom, but they are engineered rather than fixed by nature, so parameters can be designed.

3. Why It Happens

Coherence is the central constraint. The qubit loses its state through several channels: energy relaxation to the environment, dephasing from noise in control electronics and materials defects, and unwanted coupling to neighbouring circuits.

Materials have turned out to matter more than expected. Microscopic defects in surface oxides and interfaces absorb energy, and a significant fraction of coherence-time improvement in recent years has come from fabrication and materials work rather than from new circuit designs.

Every control line entering the refrigerator is a channel for noise and heat. Scaling to thousands of qubits means solving wiring and heat-load problems that are mundane but decisive.

Error correction drives everything downstream. Physical qubits are corrected by spreading one logical qubit’s information across many physical ones and repeatedly measuring error syndromes, which requires fast, high-fidelity two-qubit gates and readout.

4. Real Examples

Transmon designs: a widely used variant that deliberately reduces sensitivity to charge noise at some cost in nonlinearity.

Cryogenic control electronics: moving some signal generation and amplification closer to the qubit reduces cable count and thermal noise.

Crosstalk suppression: careful layout, shielding, and tunable couplers to keep gates on one qubit from disturbing its neighbours.

Error correction demonstrations: showing that adding more physical qubits can reduce, rather than increase, the logical error rate.

5. How It Affects Us

Progress should be judged by error rates and how they scale with system size, not by qubit counts alone. A larger machine with worse error rates is not obviously better.

The hard problems have shifted from physics toward engineering: packaging, wiring, calibration, control software, and yield in fabrication.

Useful applications, when they arrive, will likely be narrow — specific classes of simulation, optimisation or chemistry problems — rather than general-purpose speedups.

Claims about breaking encryption relate to running specific algorithms on fault-tolerant machines with qubit counts far beyond anything available now, and should be read accordingly.

6. Key Takeaways

  • A Josephson junction gives the circuit unevenly spaced energy levels, which is what makes two states individually addressable.
  • Millikelvin cooling keeps thermal noise well below the qubit’s level spacing, preserving its state.
  • Coherence is limited by materials defects, control noise and unwanted coupling — hence the emphasis on fabrication.
  • Error correction requires many physical qubits per logical qubit and pushes hard requirements onto gate fidelity.
  • Scaling bottlenecks are increasingly engineering and packaging rather than fundamental physics.