Google publish experimental proof of a major idea in quantum error-correction: Below a certain error rate, increasing the surface code size leads to exponential reductions in logical error rates. A new 105-qubit processor, Willow, is used.

Oxford Ionics prepare entangled, two-qubit states with 99.97% fidelity and implement single-qubit gates with 99.999% fidelity on a trapped-ion system.

Harvard’s Lukin group and QuEra announce a set of experimental advances, including simulating up to 48 logical qubits using error codes of varying sizes. The neutral atom system includes 280 physical qubits and logical qubits experience an improvement in error rates.

Atom Computing and IBM each announce new quantum processors having over 1000 qubits and based on neutral atom and superconducting qubits, respectively.

Single-qubit and two-qubit gates are performed on logical qubits each made up of seven physical, trapped-ion qubits. The gates include a CNOT and T gate.

Error correction is performed in real-time on a single logical qubit encoded in 10 trapped-ion qubits, with each cycle including multiple gates, measurements and processing of measurement outcomes.

IBM announces a 127-qubit processor. The processor is available for testing via IBM’s online service.

A programmable array of 256 qubits is used to simulate quantum phases of matter. The system is based on neutral atom technology.

Google simulates chains of hydrogen atoms using their Sycamore processor. The simulations involve many times more gates and twice as many qubits as any previously reported simulations on quantum computers.

Google announce the first demonstration of quantum supremacy, using a new 54-qubit processor, “Sycamore”. The experiment involves running increasingly large, random circuits on the quantum processor until they become infeasible to simulate classically.

IBM introduces their 20-qubit machine, Quantum System One, which is the first commercially available quantum computer. Dilution refrigerator included.

John Preskill coins NISQ (noisy intermediate-scale quantum) to describe quantum computing systems under 100 qubits and without error correction.

Google unveils Bristlecone, a 72-qubit processor based on superconducting qubits. Bristlecone is meant to provide a platform for Google to attempt quantum advantage but proves difficult and is downsized. Around this time, Intel announces a 49-qubit chip.

Google demonstrates a 9 qubit superconducting quantum computer and uses it to simulate molecular hydrogen. IBM unveils a 17 qubit computer, also based on superconducting qubits.

NIST researchers create the first universal, programmable quantum processor. Using a two-qubit trapped-ion system, the researchers are able to implement arbitrary two-qubit operations.

Harrow, Hassidim and Lloyd describe a quantum algorithm for finding approximate solutions to sparse systems of linear equations. The quantum algorithm is exponentially faster than the best classical alternatives and expands the possible applications of quantum computers beyond materials simulation and cryptanalysis.

Satoshi Nakamoto publishes the Bitcoin whitepaper

NMR techniques are extended to a 12 qubit system.

Researchers at Georgia Tech and Harvard implement a quantum memory, using photons to transfer quantum information between cold atoms.

Researchers create a trapped-ion computer using a single Calcium ion and perform the Deutsch-Josza algorithm using two qubits encoded in the ion’s electronic and vibrational states. The same lab demonstrates a CNOT gate using a trapped-ion computer later that year.

Using optical tweezers, single atoms are trapped and arranged in 1D and 2D arrays. Each atom can be addressed and measured individually, paving the way for neutral atom quantum computing.

A new architecture for quantum computing is proposed, showing how to build photon-based quantum computers using single-photon sources, linear optical elements and photodetectors.

Researchers at IBM Almaden and Stanford factor 15 with an NMR computer using seven active nuclear spins as qubits, implementing Shor’s factoring algorithm for the first time.

Linden and Popescu show that entanglement is necessary for a broad set of quantum operations, which weakens the case for NMR as a scalable architecture for quantum computers.

Researchers at IBM Almaden and Stanford University implement the order-finding subroutine of Shor’s factoring algorithm on a small quantum device, indicating progress toward quantum-based cryptanalysis.

Scientists at Los Alamos employ nuclear magnetic resonance to operate a seven-qubit system, enabling quantum gate operations through molecular spin control, though they don’t use it to run a quantum algorithm.

Michael Nielsen and Isaac Chuang release their textbook ‘Quantum Computation and Quantum Information.’ Also known as ‘Mike and Ike.’

NMR is used to create a 5 qubit computer. The Deutsch-Josza algorithm is run on 4 qubits.

A Japanese team shows that a qubit can be constructed and controlled using a small superconducting circuit and a Josephson junction.

The Gottesman-Knill theorem shows that certain quantum computations can be efficiently simulated on classical computers. Such computations are easier to simulate and test as a result, but also less likely to be the basis of practical applications for quantum computers.

Jones and Mosca, Oxford, use an NMR quantum computer to implement a 2 qubit solution to Deutsch’s problem. It’s the first experimental realisation of a quantum algorithm. Isaac Chuang’s team report similar results very soon after, including searching over 4 states with Grover’s algorithm.

A team based at Los Alamos implements the first 3 qubit quantum computer, using trichloroethylene molecules. The computer is used to demonstrate a 3-bit error correcting code.

Kitaev proposes a 2D error-correcting code which forms the basis of quantum error correction and is later called the surface code.

Two teams demonstrate quantum logic gates and small qubit systems using techniques from nuclear magnetic resonance spectroscopy. These major experimental results establish NMR as a testbed for early quantum computing.

As part of a review of the field, David DiVincenzo sets out 5 minimum requirements for realising a quantum computer. These criteria become a major reference point when assessing quantum computing architectures.

Lou Grover describes a quantum algorithm for searching databases that is faster than classical algorithms by a quadratic factor, demonstrating another potential application of quantum computers.

Shor and Calderbank propose one of the first error correcting codes for quantum computers, showing how to protect computations from decoherence.

Scientists at NIST perform one of the first experimental demonstrations of a fundamental quantum logic gate, the CNOT gate. Laser cooled, trapped Beryllium ions act as the qubits.

Chuang and Yamamoto show how the Deutsch-Josza algorithm can be implemented with techniques from optics, providing a concrete route to realising a photonic quantum computer.

Peter Shor describes two groundbreaking algorithms which show how quantum computers can solve factoring problems and discrete log problems exponentially faster than classical computers. Quantum computers suddenly have a major new application: breaking public key encryption.

Simon sets out to show that quantum computers cannot achieve exponential speedups… and proves himself wrong. He proposes an eponymous problem and finds a quantum algorithm that is exponentially faster than classical solutions. Despite being rejected from the conference he submitted his result to, Simons inspires Peter Shor to pursue this line of research.

Bernstein and Vazirani expand on the universal quantum computer and the Deutsch-Josza algorithm to introduce BQP, a computational complexity class that distinguishes quantum computing capabilities from those of probabilistic and classical computers.

Deutsch and Josza prove that quantum computers can perform some computations faster than classical computers, suggesting that quantum speedups might exist for problems other than simulating physical systems.

Ekert shows how to communicate securely using entanglement and quantum channels. Users can check if somebody is eavesdropping using Bell inequalities.

Yamamoto and Igeta are among the first to propose a physical implementation of a quantum computer. They encode qubits in photons and show how to perform a number of gates.

David Deutsch describes a universal quantum computer, a machine which can be programmed to simulate every finite, physical system, including all other possible quantum computers.