Quantum computer

Google-quantum-supercomputer — Quantum AI supercomputer at Google

Quantum computers can use quantum-mechanical phenomena such as superposition and entanglement to perform computation. These are being built to be able to solve certain computational problems, such as integer factorization in RSA encryption, with orders of magnitude faster than a traditional computer could.

Binary code represents information as either a 0 or 1, and these traditional von Neumann architecture computers are better at serial processing. Quantum computers are better at parallel processing because they use qubits, which are simultaneously both 0 and 1.

Qubits are formed by the quantum states of particles like electrons or photons, switching them between quantum states (representing 0, 1, or superpositions of both). This superposition of states allows a quantum computer to test every possible combination of qubits at the same time.

The Hilbert space or register of a quantum system represents all possible states that the system’s qubits can occupy. A quantum computer with 50 qubits gives a Hilbert space of size 2⁵⁰. This can be a physical limitation. A thousand qubit system could test $2^{1000}$ potential solutions simultaneously, thus vastly outperforming a conventional computer even with parallel processing across thousands of CPUs. To get a sense for the magnitude of $2^{1000}$ (which is approximately $10^{301}$ ), note that there are only about $10^{80}$ atoms in the visible universe.

As of 2024 there is no working quantum computer with 100% error-correcting and coherent qubits. Error correction is around 99.9% which by some accounts is regarded as a software level "critical mass". It is therefore an engineering problem requiring billions of dollars for superconductive cooling and energy. There may be low-cost, low-energy solutions using optics. A working quantum computer would actually prove that quantum mechanics is not flawed. An analogous problem is the race towards the world's first AGI.

These are the approaches to building a quantum computer:


Superconducting	Google/IBM	A superconducting quantum computer (like at Google and IBM) uses superconducting circuits as qubits. The qubits are tiny loops of superconducting material that can carry a current in different states simultaneously, due to quantum superposition. Microwave pulses are used to control and manipulate the states of these qubits, creating quantum gates. The qubits can be entangled (with two qubit gate fidelity), leading to quantum parallelism. The final state of the qubits is measured to get the output, collapsing their quantum states to either 0 or 1. News: 2019: Google claimed to have achieved quantum supremacy via Sycamore.
Trapped ion	Microsoft/Continuum	A trapped-ion quantum computer uses charged atoms (ions) to perform quantum computations. Each ion represents a qubit, and are held in place using electromagnetic fields in a vacuum chamber. Lasers are used to change the state of the ions and make them interact, allowing quantum gates and operations to be performed. Measurements are done by shining a laser on the ions, causing them to fluoresce (light up) or not, depending on their state, returning their qubit states.
Neutral atom	Harvard/QuEra	A neutral atom quantum computer uses individual, electrically neutral atoms as qubits. Lasers create "optical traps" or lattices of light, which hold each atom in a specific location. Laser pulses are used to change the energy states of the atoms, effectively switching them between quantum states. Controlling the atoms' positions and interactions creates entangled pairs. The final quantum states of the atoms are read by observing how they interact with light.
Majorana	Microsoft	A Majorana quantum computer aims to use Majorana fermions as qubits. Majorana fermions exist in pairs, and each pair together forms one qubit. They can "remember" their quantum state even in noisy environments, which helps prevent errors. They are manipulated by braiding, where the positions of the particles are swapped to create quantum gates. News: February 19, 2025: Announcement and Microsoft CEO statement Claims: The world’s first Quantum Processing Unit (QPU) powered by a Topological Core. An entirely new state of matter called topoconductors. Able to create Majorana particles. A path to a million qubits.
Photonic	LightSolver Jiuzhang	Optical computers could eventually represent the pinnacle of quantum computing, with superior speed, low energy costs, and stable qubits using light pulses or photons. News: 3 December 2020: Jiuzhang claims it has achieved quantum supremacy
Quantum dot	Hypothetical	Qubits are made using quantum dots. A single electron is trapped inside a quantum dot. Its spin state represents 0 or 1, just like a classical bit. Using magnetic or electric fields, the electron’s spin is manipulated, enabling quantum superposition and entanglement. They would be more stable and scalable than other qubit types and work well with existing semiconductor technology. Potential for large-scale quantum computers in the future.
Moleculartronic	Hypothetical	A quantum computer using nanotechnology is a moleculartronic computer.

Time crystals are used to stabilize qubits. They are resistant to entropy and can be used for quantum memory. A time crystal was observed by Google.

Applications suited to a high degree of parallelism and concurrency:

Quantum entanglement and modelling for quantum teleportation systems.
Computational chemistry, for example a new drug R&D program is trial and error, iterative, time-consuming and expensive.
Particle physics workflows such as needed for colliders; quantum chromodynamics and quantum physics.
Molecular modeling such as in chemical reactions that are quantum in nature, forming entangled quantum superposition states.
Global weather prediction and modelling. 30 percent of world GDP is directly or indirectly affected by weather, impacting food production, transportation and trade. Another benefit is giving ourselves time to take cover from disasters.
Artificial Intelligence and Machine Learning, based on calculating the probabilities for many possible choices.
Cybersecurity and cryptography. New quantum-resistant cryptography methods would be needed that would resist quantum computers able to crack the old prime factorization methods.
Financial Modelling and stock markets. Randomness inherent to quantum computers is congruent to the stochastic nature of financial markets. Investors often wish to evaluate the distribution of outcomes under an extremely large number of scenarios.
Combinatorics and set theory
Chaos Theory
Simulating the electronic structure of a small molecule

Although no current quantum computers are powerful enough to do so against any real implementations, a vulnerability for traditional encryption such as that used in cryptocurrency, and the most popular cryptographic algorithm, as of 2022, AES, is that quantum computers will be able to crack them. Math problems, like the ones encrypting our sensitive data today, that, for certain algorithms, would take a traditional supercomputer millions of years to solve could be solved by theoretical future quantum computers in downwards of 48 hours. In a cryptographic attack known as "capture now, decrypt later", people and governments are harvesting encrypted data today to, in theory, decrypt later with quantum computers. In the near future once quantum computers are mainstream, they will be used to breach our current non-resistant cryptographic defenses.

Two ways of circumventing this problem would be to create traditional encryption algorithms that are quantum resistant, Post-quantum cryptography, or adopt quantum encryption algorithms, quantum cryptography, that would be impossible for traditional or quantum computers to crack. Quantropi claims to be one of the world's first companies to offer the latter method with their non-photonic quantum key distribution that works over the cloud - so clients are able to use existing infrastructure and the internet to encrypt their sensitive information into quantum states of entropy. Quantropi’s QEEP symmetric encryption ensures "uncertainty" to attackers, rendering data uninterpretable forever. They also provide a quantum-secure layer to AES encryption.

Shor's algorithm is a quantum algorithm for finding the prime factors of an integer. If a quantum computer with a sufficient number of qubits could operate without succumbing to quantum noise and other quantum-decoherence phenomena, then Shor's algorithm could be used to break public-key cryptography schemes, such as the RSA scheme. Breaking traditional cryptography, as discussed above, would have a fundamental effect on IT security systems around the world (for example online banking) and working quantum cryptography systems would have to be adopted.

Quantum Supremacy[]

This has been demonstrated since 2019 using the following premises:

state a computational problem that is computationally expensive for a classical computer using all known algorithms and techniques
stated problems have so far not been scientifically useful or practical or reusable for real-world applications
stated problem is much easier and less computationally expensive for a quantum computer
demonstrate a vast speed-up in computational time using a quantum computer

Quantum processors that have demonstrated quantum supremacy:

Google's Sycamore processor - 53 qubits
China's photonic Jiuzhang - 53 qubits
China's Zuchongzhi processor - 66 qubits (link)
2021: IBM's Eagle chip has 127 qubits
2023: IBM's Osprey Processor has 433 qubits
IBM Condor at 1121 qubits released in 2023
Atom Computing at 1180 qubits in October 2023

However, useful quantum supremacy has not been achieved. If you wanted to factor a 20-digit semiprime number, a quantum computer cannot solve this problem at all. A classical computer can do this in milliseconds.

Achieving useful quantum supremacy would enable us to:

make high-performance quantum chemistry and quantum physics calculations
solve a problem that’s actually relevant to the real world, such as the double slit experiment in the Heisenberg Uncertainty Principle
replace all classical computers trying to solve computationally expensive problems with superior quantum computers
run Shor’s algorithm for arbitrarily large numbers

A quantum computer would not replace classical serial operations such as those found on operating systems. A PC or smartphone would not be replaced. A supercomputer that is taking too long to solve a problem will be replaced by a quantum computer that will solve the same problem in a drastically faster time.

Quantum Advantage[]

The quantum supremacy and quantum advantage definitions have become blurred and almost synonymous. Current preference is for quantum advantage, since "supremacy" has a negative connotation.