A different kind of machine, not a faster one
A quantum computer is not simply a quicker version of the computers we use today. It follows different rules and is useful only for certain kinds of problems. For most everyday jobs, such as browsing, writing documents, streaming video or running apps, an ordinary computer is and will remain the better tool.
The basic difference is in how information is stored. A normal computer uses bits, each of which is either 0 or 1. A quantum computer uses qubits, which are built from physical systems that follow the rules of quantum physics.
What a qubit is
A qubit can be placed in a state that is described as a combination of 0 and 1 at once, known as superposition. A common but imperfect picture is a spinning coin that is neither heads nor tails until it lands. When the qubit is measured, it gives a definite 0 or 1, with probabilities set by its prior state.
A second feature is entanglement, in which the states of several qubits are linked, so that the group has to be described together rather than one by one. Quantum algorithms are designed to arrange these linked states so that wrong answers cancel out and the right ones become more likely when the result is measured.
It is easy to oversimplify this. A quantum computer does not try every answer at the same time and then pick the best one. The skill lies in designing an algorithm in which the interference between possibilities favours the correct outcome, and that is only possible for particular problems.
Why building one is hard
Qubits are fragile. Heat, vibration, stray electromagnetic signals and even the act of measuring can disturb them and introduce errors. Many designs need extreme cooling or very carefully controlled conditions. Today’s machines make errors far more often than ordinary processors do.
Researchers work on error correction, which combines many physical qubits to act as one more reliable “logical” qubit. This is expected to need large overheads, and building machines big and stable enough to run long, useful calculations remains an engineering challenge. Forecasts about when this will happen vary widely, and nobody can state a reliable date.
What quantum computers may be good at
Researchers expect advantages for a limited set of problem types.
- Simulating nature. Molecules and materials are themselves quantum systems, so quantum computers may help model them, which could eventually assist work in chemistry and materials science.
- Certain optimisation and search tasks. Some methods are believed to offer speedups, though the size of the benefit often depends on the exact problem, and for many practical cases it is not yet proven.
- Factoring large numbers. A well-known algorithm shows that a sufficiently large, error-corrected quantum computer could factor big numbers far faster than known methods on ordinary computers.
For most business problems, there is no demonstrated quantum advantage today. Some experiments have shown a quantum device completing a specially chosen task that is hard for classical computers, but those tasks were generally not useful in themselves. Claims should be read with that distinction in mind.
What they cannot do
- They will not replace everyday computers or make general software run faster.
- They do not solve every hard problem. Some problems are believed to stay hard even for quantum machines.
- They are not a route to science-fiction outcomes such as reading minds or predicting the future.
- Current devices are small and noisy, so many proposed applications are still theoretical.
It is sometimes suggested that quantum computing will transform machine learning. That is an active research topic, but practical benefits have not been established, and most machine learning today runs well on conventional chips.
The effect on encryption
The most discussed practical issue is security. Much of the encryption that protects websites, payments and messages relies on mathematical problems, such as factoring large numbers, that are very hard for ordinary computers. The factoring algorithm mentioned above means that a large enough quantum computer could, in principle, break some widely used public-key methods. Other kinds of encryption, such as the symmetric methods used to protect stored and transmitted data once a connection is set up, are thought to be affected less, and longer keys are generally considered a sensible response.
No quantum computer today is believed to be capable of this, and experts disagree about when, or if, one will be. Even so, there is a concern sometimes called “harvest now, decrypt later”: someone could record encrypted data today and try to read it years later if the technology arrives. For information that must stay secret for a long time, that is a reason to plan ahead.
Standards bodies and security researchers have been developing new methods, often called post-quantum cryptography, designed to resist both ordinary and quantum attacks. Software and device makers are gradually adopting them. For ordinary users, the practical steps are modest: install software updates promptly so that stronger methods arrive automatically, and keep using established tools such as end-to-end encryption and device encryption. There is no need for individuals to do anything special.
Other things people ask about
Digital currencies and blockchain technology rely on cryptographic methods too, so they are part of the same discussion, and developers have considered how to adapt them. The details vary by system.
How to read the headlines
News about quantum computing often mixes real laboratory progress with promotional language. Useful questions to ask are what specific problem was solved, how it compares with the best ordinary computer methods, and whether the result is useful outside a demonstration. Genuine progress is being made, but a broadly useful quantum computer remains a goal rather than a product, and its eventual effects are uncertain.