Quantum technologies, explained

The basics in five ideas

• Qubits: A bit is 0 or 1. A qubit can be in a blend of both — a superposition — described by two numbers that set how likely each outcome is.
• Entanglement: Qubits can be linked so that their outcomes are correlated more strongly than anything classical allows. n entangled qubits are described by 2ⁿ numbers.
• Interference: An algorithm steers those numbers so wrong answers cancel out and right ones add up. Speed-ups exist only for problems with that kind of structure.
• Measurement: Reading a qubit gives a plain 0 or 1 and destroys the superposition, so a program is run many times and the results are counted.
• Noise & error correction: Qubits lose their state through tiny disturbances. Error correction spreads one reliable “logical” qubit over many physical ones — the main race in the field today.

Ways to build a qubit

Superconducting circuits

Tiny electrical circuits on a chip, made of metals that lose all resistance when very cold. A Josephson junction makes the circuit behave like an artificial atom with two usable energy levels; microwave pulses control it.

Runs at: About 10–20 millikelvin in a dilution refrigerator

Strengths:

• Very fast operations (tens of nanoseconds)
• Built with chip-fabrication techniques
• The most mature software and cloud access

Challenges:

• Qubits forget their state quickly
• Each qubit talks mostly to its neighbours
• Wiring and cooling thousands of qubits

Who builds it: IBM, Google, Rigetti, IQM, Alice & Bob, AWS

Trapped ions

Single charged atoms held in place by electric fields inside a vacuum chamber. Lasers or microwaves change their internal state, and the ions’ shared motion is used to entangle them.

Runs at: Ultra-high vacuum; the trap is often cooled, the ions are laser-cooled

Strengths:

• Every ion is identical by nature
• Among the most accurate operations of any platform
• Any qubit can interact with any other in a chain

Challenges:

• Slower operations (microseconds or more)
• Long chains get harder to control, so designs move ions between zones
• Complex laser and vacuum hardware

Who builds it: Quantinuum, IonQ, AQT, eleQtron

Neutral atoms

Uncharged atoms held by tightly focused laser beams (“optical tweezers”) in large 2D or 3D arrays. Exciting them to a big, sticky “Rydberg” state lets neighbours interact.

Runs at: Vacuum; atoms laser-cooled to microkelvin, the apparatus itself near room temperature

Strengths:

• Hundreds to thousands of qubits in one array
• Atoms can be moved to change who talks to whom
• Also used as analog simulators

Challenges:

• Atoms get lost and must be reloaded
• Slower cycles than chips
• Operation accuracy is still catching up

Who builds it: QuEra, Pasqal, Atom Computing, Infleqtion, planqc

Photons

Qubits carried by particles of light — in their path, timing or polarisation — guided through optical chips or fibre. Because photons barely interact, many designs entangle them through clever measurements instead.

Runs at: Mostly room temperature; the single-photon detectors are usually cryogenic

Strengths:

• Light travels well, so it fits naturally with networks
• Can reuse silicon-photonics factories
• Little noise from the environment

Challenges:

• Every lost photon is a lost qubit
• Making photons interact on demand is hard
• Needs very large numbers of components

Who builds it: PsiQuantum, Xanadu, Quandela, ORCA Computing, QuiX Quantum

Spin qubits in silicon

The spin of a single electron (or nucleus) trapped in a “quantum dot” — a nanoscale cage built into a silicon chip much like a transistor. Magnetic or electric pulses flip the spin.

Runs at: Below about 1 kelvin — warmer than superconducting qubits

Strengths:

• Qubits are tiny, so millions could fit on a chip
• Could use existing semiconductor factories
• Control electronics could sit close to the qubits

Challenges:

• Small differences between dots from fabrication
• Systems are still small
• Wiring a dense array

Who builds it: Intel, Diraq, Quantum Motion, Equal1, SemiQon

Topological qubits

Information stored in the shape of exotic states (Majorana modes) spread across a nanowire rather than in one particle, which in theory protects it from local noise.

Runs at: Millikelvin, in superconductor–semiconductor devices

Strengths:

• Errors would be rare by design
• Could need far fewer physical qubits per logical qubit

Challenges:

• The basic building block is still being demonstrated
• Published claims are debated by researchers

Early stage: no topological qubit has yet run a computation.

Who builds it: Microsoft, Nokia Bell Labs

Quantum annealers

A different kind of machine: instead of running a sequence of gates, it encodes an optimisation problem as an energy landscape and lets thousands of superconducting qubits settle into a low-energy answer.

Runs at: Millikelvin, like other superconducting chips

Strengths:

• Thousands of qubits available today
• A direct fit for scheduling and routing problems

Challenges:

• Not a general-purpose quantum computer
• A clear advantage over the best classical methods is still argued

Annealers solve one type of problem; the other platforms aim at universal, gate-based computers.

Who builds it: D-Wave

Side by side

Beyond computing

Quantum communication

Uses single photons and entanglement to send information. The best-known use is quantum key distribution (QKD): two parties share an encryption key, and any eavesdropper disturbs the photons and gets noticed. Longer term, quantum networks would link quantum computers and sensors.

Examples: ID Quantique, Toshiba, Qunnect, Aliro

Quantum sensing

The same fragility that troubles computers makes quantum systems superb detectors. Atomic clocks, cold-atom gravimeters, and diamond NV-center magnetometers measure time, gravity and magnetic fields with extreme precision — the most commercially mature quantum technology.

Examples: Qnami, Exail, Q-CTRL, Infleqtion