The Real Workings of a Quantum Computer

A Quantum Computer Does Not Try All Answers at Once

The most common analogy for a quantum computer is wrong. It is a seductive image: a processor that checks every possible solution to a problem simultaneously, like a key that can try every lock in a hallway at the exact same instant. This mental model suggests a kind of brute-force parallelism on an unimaginable scale, and it is the source of much of the popular excitement and science-fiction speculation surrounding the technology. It is also deeply misleading.

If a quantum computer worked this way, measuring the result would be like reaching into a bag containing all possible answers and pulling one out at random. The correct answer would be in there, but so would every wrong one. The probability of finding the right solution would be vanishingly small for any problem of meaningful complexity. The machine would be useless. The real source of quantum computational power is stranger, more subtle, and rooted in the wave-like nature of quantum reality.

The Qubit Holds a Spectrum of Possibilities

A classical computer uses bits. A bit is a switch. It is either off or on, a 0 or a 1. Its state is definite and exclusive. A quantum computer uses a quantum bit, or qubit. A qubit is not simply a 0 and a 1 at the same time. Instead, it exists in a state of superposition, which is best described not as a coin spinning in the air, but as a vector pointing in a specific direction within a conceptual space.

This direction is defined by two numbers called probability amplitudes. One amplitude is associated with the outcome 0, the other with the outcome 1. When the qubit is measured, its state collapses to either 0 or 1 with a probability determined by these amplitudes. Critically, and unlike classical probabilities, these amplitudes can be positive, negative, or even complex numbers. This single fact is the gateway to understanding quantum computation. A negative amplitude has no classical analogue, yet it is essential. This behaviour is a direct consequence of the rules governing the subatomic world. The principles of <a href="/what-is-quantum-mechanics">Quantum Mechanics in Four Slits</a> govern this reality, which has no parallel in our everyday experience.

Interference Is the Source of Quantum Power

Because amplitudes can be negative, they can cancel each other out. This is called interference, the same phenomenon seen when two waves on a pond meet. If their crests align, they form a bigger wave (constructive interference). If a crest meets a trough, they flatten out and disappear (destructive interference). A quantum algorithm is a carefully choreographed sequence of operations—quantum gates—designed to manipulate the amplitudes of a system of qubits. The goal is not to explore all paths. The goal is to make the computational paths leading to incorrect answers interfere destructively, while making the paths leading to the correct answer interfere constructively.

The computer is not trying all keys in the hallway at once. It is performing a single, complex evolution of a quantum state. This evolution is engineered so that, at the moment of measurement, the amplitudes for all the wrong keys have cancelled themselves to zero, while the amplitude for the one right key has been amplified close to one. The physicist David Deutsch first formalised the idea of a universal quantum computer in the 1980s, showing that such a machine could, in principle, efficiently simulate any physical process that occurs in nature. This hinted at a power beyond classical machines.

Key Algorithms Reveal the True Nature of the Speedup

The potential of this interference-based computation is best seen through its most famous algorithms. They do not speed up every task. They only work on problems with a specific mathematical structure that can be mapped onto a quantum interference pattern.

Building a Quantum Computer Is Immensely Difficult

The same quantum properties that give these machines their potential also make them exquisitely fragile. The primary obstacle is a phenomenon called decoherence. The slightest interaction with the outside world—a stray magnetic field, a tiny temperature fluctuation—can disturb the delicate superposition of the qubits, destroying the computation in a fraction of a second. This is often called noise.

To combat this, researchers are developing quantum error correction. The concept involves encoding the information of a single, ideal 'logical qubit' across many faulty 'physical qubits'. By constantly measuring the state of these physical qubits for errors (in a clever way that does not destroy the underlying computation) and correcting them, a robust logical qubit can be maintained. The overhead is enormous. Current estimates suggest thousands, or even millions, of physical qubits might be needed to create a single, fully error-corrected logical qubit. Today's machines have at most a few hundred physical qubits, and they are all noisy.

There is no consensus on the best way to build a physical qubit. Several competing technologies are being pursued: superconducting circuits cooled to near absolute zero; individual charged atoms (ions) held in place by electromagnetic fields; photons, which are particles of light, manipulated by mirrors and beam-splitters; and neutral atoms trapped by lasers. Each has distinct advantages in qubit stability, connectivity, and gate speed, and each has its own severe engineering challenges. Furthermore, not all quantum computers are the universal, gate-based machines described above. Some, like quantum annealers, are designed to find the lowest energy state of a system, making them useful for specific optimisation problems but not for tasks like running Shor's algorithm.

Progress Is Real, but Advantage Remains Contested

In 2019, a team at Google announced they had performed a calculation on their 'Sycamore' quantum processor that would be effectively impossible for even the most powerful classical supercomputer. They claimed to have achieved 'quantum supremacy'. The task itself, which involved sampling the output of a random quantum circuit, had no practical purpose. It was a benchmark. The claim was immediately disputed by competitors like IBM, who argued that with better algorithms and enough resources, a classical supercomputer could in fact perform the task, albeit much more slowly. This debate highlights a crucial point: the frontier of quantum advantage is a moving target, not a fixed line.

The threat to cryptography, however distant, is being taken seriously. A global effort is underway to develop and standardise 'post-quantum cryptography' (PQC)—new encryption methods that can run on today's classical computers but are believed to be resistant to attack by both classical and quantum machines. The U.S. National Institute of Standards and Technology (NIST) has already selected a first suite of PQC algorithms for standardisation. The transition to these new standards will be a slow and complex undertaking over the coming decade.

The Markers of Genuine Progress

Hype will continue to cloud the field. True progress will be marked by concrete, verifiable achievements. Watch for the first demonstration of a fault-tolerant logical qubit that is provably more stable and useful than its constituent physical qubits. Watch for a quantum simulation of a molecule that yields a discovery—like a new catalyst—unobtainable by classical means. Watch for the factoring of a cryptographically relevant (though still small) number with Shor's algorithm on a real device. These are the milestones that will signal the dawn of useful quantum computation. Until then, a healthy, informed scepticism is the correct approach.