quantum computer
A computer that uses the laws of quantum physics to solve certain special tasks much faster than current machines.
In plain terms
A conventional computer works with bits: every bit is either 0 or 1, like a light switch. The building block of a quantum computer is the quantum bit, or qubit, which during calculation can also be in a peculiar mixture of both. It is a bit like a spinning coin: while spinning, it is neither heads nor tails; it only settles when you slap it. The analogy is flawed because qubits can also link with each other in a way coins cannot, and the computer's power stems precisely from this.
This machine will not be faster at everything. It is more of a specialized tool than a stronger engine: it promises huge advantages in some task types, but none in most everyday work. The most certain areas are simulating the behavior of molecules and materials, and breaking some of the currently used encryptions. Researchers also hope for benefits in optimization tasks, such as route or schedule planning, but the advantage there is not yet proven.
This is only loosely related to artificial intelligence. Today's chatbots and AI agents run on traditional chips, not quantum computers. Researchers are investigating whether quantum computers might one day assist AI, but this remains an open question for now.
How it works
- The machine uses qubits, which are realized, for example, by superconducting circuits, trapped ions, laser-trapped neutral atoms, or photons. Superconducting qubits operate in a refrigerator colder than outer space.
- Qubits are put into superposition, meaning they simultaneously carry the possibility of 0 and 1, and they are entangled, so their states are interconnected.
- A quantum algorithm uses a series of operations to shape this shared state so that the probability of the correct answer is amplified while that of incorrect answers is cancelled out, much like waves reinforcing or cancelling each other.
- At the end, the qubits are measured, at which point they become simple 0s and 1s. Since the result is probabilistic, the calculation is usually repeated many times.
- Since qubits are extremely sensitive to heat and noise, many physical qubits must be combined into a single reliable, so-called logical qubit. This is error correction, the field's greatest engineering challenge.
An everyday example
You will likely never work directly with a quantum computer yourself, but you will encounter its impact. Your online banking, browser, and messaging app are gradually switching to new encryption in the background that, according to current knowledge, even a future quantum computer cannot break. The only thing you might notice is that the app updates.
Who is affected
A common misconception
Limits and risks
- Today's machines are noisy and error-prone, so in practical, business-useful tasks they still do not reliably surpass traditional supercomputers.
- The number of qubits alone tells little: their quality and error rate matter just as much, so news about records should be read with caution.
- The encryption risk is real but not imminent: breaking the currently widespread keys would require a much larger, error-corrected machine than exists today, and no one knows for sure when such a machine will be ready.
- Machines are expensive and require delicate conditions, so they will not become home or office devices in the foreseeable future.
Similar, but not the same
- supercomputer – A supercomputer connects many traditional processors, so it does the same thing as your laptop, just on a much larger scale, whereas a quantum computer calculates based on different physical principles.
- AI-chip (GPU) – Today's AI models run on traditional graphics processors that work with bits; these have nothing to do with qubits.
- post-quantum cryptography – This does not run on a quantum computer: it is encryption written for traditional computers, designed so that even a quantum computer cannot break it.
Related concepts
computing, quantum technology, computing infrastructure
qubit, kvantumalgoritmus, quantum error correction, quantum supremacy
supercomputer, post-quantum cryptography, encryption, GPU, data center
A brief history
The idea was raised in the early 1980s by physicists including Paul Benioff and Richard Feynman; Feynman argued that quantum nature could best be simulated with a quantum machine. In 1994, Peter Shor published an algorithm with which a sufficiently large quantum computer could break widely used public-key cryptosystems, such as RSA. In 2019, Google was the first to announce that its quantum processor had outperformed supercomputers on an artificial task with no practical use. This result was later contested, because with improved classical methods the task could be solved much faster than expected. In 2024, the US National Institute of Standards and Technology (NIST) issued the first post-quantum cryptography standards.
Where things stand
By the end of 2026, quantum computers are available in research laboratories and as cloud services, with remarkable progress in error correction. They have not yet brought widespread, everyday practical benefits. However, the replacement of encryption with quantum-resistant alternatives is already underway based on NIST standards issued in 2024.
Also known as
quantum computer · quantum computer · quantum computer · kvantumkomputer · quantum computing · kvantuminformatika · quantum computer · quantum computing
Sources
This entry was machine-edited and has undergone machine expert review; the “In the news” section updates automatically from Agent Age stories.