US research and data company Lux Research has, in its most recent report, 'Preparing for Quantum Computing', cautioned that quantum computing is still at least a decade away from being a useful, reliable technology for most businesses. While quantum computing promises to be able to solve complex problems beyond the capacity of conventional supercomputers, it still faces problems that limit its capabilities in practice, and it is unclear if all of these will be overcome during the next decade.
“Today’s supercomputers tackle difficult problems including weather modelling, genomic analysis, and computational fluid dynamics, but even the best supercomputers will always be limited in specific areas,” pointed out Lux senior research associate and report lead author Lewie Roberts. “They’re still unable to handle some important problems like chemical product design, protein folding, or supply chain optimisation. It’s our belief that quantum computing will one day enable multiple industries to address some of these key problems, moving past today’s barriers and enabling further innovation.”
All computers store and manipulate information. Conventional computers store information in binary 0 and 1 form, and each binary digit is called a bit; these individual bits are manipulated by the computer. Quantum computers operate on a completely different basis. They use the particle physics phenomena discovered through quantum mechanics such as superposition and entanglement (as well as the long-known phenomenon of interference). The basic unit of information in a quantum computer is a qubit (from quantum bit); qubits are not binary digits but represent atoms, ions, photons or electrons, plus their various control devices, which operate together as memory systems and processing systems.
Superposition is a combination of states that are normally described separately – if a musician plays two musical notes at once, that is a superposition of those notes. While a bit must be either a 0 or a 1, because of superposition a qubit can simultaneously be 0 and 1, and every fraction in-between. Entanglement is a phenomenon in quantum mechanics in which the quantum states of two or more particles must be described in relation to each other, not separately, even if they are physically separate. The consequence of these phenomena is that, while conventional computers have to process different problems sequentially, quantum computers can process different problems simultaneously.
However, at the moment, only small scale quantum computers exist, because of the complex and difficult engineering required to create them. Furthermore, qubits are inherently unstable, which impairs the accuracy of any computation which makes use of them. Thus, quantum computing is currently being used mainly for problems that do not have clearly defined answers, such as machine learning, or the simulation of quantum systems. Scientists and engineers developing quantum computing hardware are seeking to increase the stability of qubits. But Lux believes that, in the end, they will have to create fault-tolerant quantum computers able to correct the errors created by qubit instability, and that such fault-tolerant quantum computers are unlikely to be developed during the next decade. For most businesses, “[q]uantum computing is not currently providing business value that could not be achieved with today’s existing computers, and it’s not clear when it will,” observed Roberts.