Quantum Computing

Every normal computer calculates using switches that are either off or on. This is written as zero or one, and everything else — text, images, videos — is built on top of that. A quantum computer instead uses tiny particles, such as individual atoms or particles of light. Such particles can exist in an intermediate state that contains both zero and one at once. Only when a measurement is taken is it decided what the outcome will be. It is precisely this strange behavior of nature that researchers want to exploit in order to solve certain computational tasks faster.

What a quantum computer could crack

There are tasks that even the largest supercomputer of today fails at. One example is breaking down very large numbers into their prime factors. It is precisely this difficulty that protects the majority of encryption on the internet today. A sufficiently large quantum computer could perform this calculation in a manageable amount of time. That is why cryptographers are already working on new methods that would remain secure even then.

More interesting for the economy is another area: the simulation of molecules. Chemical reactions themselves follow the laws of quantum physics, and classical computers have to approximate them laboriously. A quantum computer could model them more directly. Possible applications would include new medications, better batteries, or more efficient fertilizers.

An important distinction: a quantum computer is not a faster laptop. For word processing, video games, or training today’s AI models, it offers nothing. It is a specialized tool for a handful of problem types. Anyone who reads about a general speedup of all computers has fallen for a widespread misconception.

Qubits, superposition, and the problem of susceptibility to interference

The computing unit is called a qubit, short for quantum bit. A qubit can represent zero and one at the same time, which is called superposition. Multiple qubits can also be entangled: they then behave like a single system, no matter how far apart they are. With 300 entangled qubits, one could theoretically represent more states simultaneously than there are atoms in the visible universe.

A comparison helps: a classical computer tries one path after another in a maze. A quantum computer, so to speak, lays itself over all paths at once. The trick is to construct the calculation so that wrong paths cancel each other out and the correct one remains. This art is called a quantum algorithm, and so far only a few of them are known.

The catch is sensitivity. Even minimal heat, vibration, or radiation destroys the quantum state. That is why many systems operate at temperatures close to absolute zero, i.e., below minus 273 degrees Celsius. Nevertheless, the machines constantly make errors. Error correction is therefore needed, in which many physical qubits are combined into a single reliable logical qubit — depending on the method, a thousand or more.

From the lab to the headlines

You can practically not buy a quantum computer. Google, IBM, Microsoft, and a number of start-ups operate their devices in laboratories and rent out computing time over the internet. Anyone who wants to can try out small programs this way, often even for free. So far, productive use is almost exclusively limited to research.

In the news, the term usually appears in two contexts. On one hand, in record-breaking announcements: new chips with more qubits or better error correction. On the other hand, in stock prices, since publicly traded quantum companies react strongly to such announcements. Caution is advised when only the number of qubits is mentioned. Without information on the error rate, it says little about actual performance.

A frequent buzzword is quantum supremacy. This refers to a quantum machine solving a task faster than any classical computer. Such demonstrations have already taken place, though with artificial tasks that have no practical use. Most experts expect it to take ten years or more before useful applications arrive in everyday life.