Already of interest is the use of quantum computers to perform quantum simulation. One example is calculating the quantum states of molecules involved in chemical energy storage, which will improve battery technology. Similarly, the details of various chemical reactions involved in the efficacy of pharmaceutical drugs are beyond the capabilities of classical computing, but may be achievable on quantum computers with relatively few (perhaps around 100) physical qubits. Research is already approaching relevance, with large-scale chemistry simulations running on 70+ qubits, and useful hybrid quantum-classical approaches such as Algorithmiq’s “Majorana Propagation”, which aims to ensure that future quantum computations will be relevant, accurate and scalable.¹¹

Other healthcare and life-science applications are also envisaged.¹² Many quantum algorithms for quantum simulations use both quantum computers and classical ones in an integrated manner. This is because only part of a very complex molecular simulation benefits from quantum algorithms; the rest of it can be performed by conventional simulation methods, including machine-learning techniques. This framework is known as hybrid quantum computing and it is considered nowadays the only path to scalability.

There is also increasing interest in using quantum computers to assist machine-learning algorithms¹³ and financial modelling.¹⁴ Similarly, logistics applications, such as routing and scheduling, may benefit from the advantages offered by quantum search algorithms. As well as mathematically rigorous algorithm design, there is scope for heuristics, where quantum-computing advantages can be achieved through trial-and-error approaches.

Researchers have long planned to integrate quantum error correction (QEC) resources into their quantum processors.¹³Theoretical research indicates that holding the bit in a network of qubits designed to detect and correct errors as they arise can significantly reduce the chances of a computation going awry. Examples include the surface code¹⁴ and the low density parity code (LDPC).¹⁵ Though this increases the required number of qubits, research in this area has uncovered ways in which this overhead can be reduced.¹⁶

To avoid the overhead of error correction, some other techniques have been effective for NISQ devices. These include the implementations of “error suppression”, which decreases the errors created in hardware operations, and “error mitigation” routines that inspect circuit operation and inject “inverted” noise to clean up qubit states while a computation is ongoing.^{17 18} Other recent error mitigation methods deal with noise fully in post-processing, that is after the readout of quantum computers, and have been proven to be optimal.^{19 20} Noise-reduction techniques applied to the classical electronics used to control the circuit have also proved to be useful. These can reconfigure the electronics’ settings and the signals applied so that the information held in qubits is better protected.

One area of imminent application is in light-activated cancer drugs.²² Quantum computation allows researchers to design drugs with molecular energy states that allow them to be activated with particular wavelengths of light, which makes the process of generating cancer-killing reactive oxygen species less harmful to the patient than can be achieved using classical means.

Near-term quantum computers could also speed up readouts from quantum sensors and machine-learning-based tasks,²³ and simulate realistic models of materials, thanks to recent understanding of measures that can reduce the resources required by several orders of magnitude.²⁴ XPRIZE Quantum Applications, a competition designed to generate algorithms that can help solve current real-world challenges, may also stimulate progress.²⁵ (GESDA is the presenting sponsor of this initiative.)