Quantum computing

How we think about quantum computing

Barclays first started to explore quantum computing in 2017. We spent a significant period of time learning about the technology until we felt had a strong internal understanding of the technology. We then concluded that quantum computing’s potential was so great that we should commit to an initial programme of research and development.

Since then, we have undertaken in-depth experimentation using the technology, to find out how it could one day be used to make improvements in the banking industry. The prize on offer here is twofold: we want to grow our banking industry’s understanding of quantum computing generally and, at the same time, give ourselves insights that could potentially be leveraged commercially in future.

Quantum computing at Barclays today

We are currently conducting experiments that focus on some very specific challenges we come across in the banking industry, which will help us learn how quantum computing could benefit particular types of work that we do. For example, our initial tests focused on how quantum computing could potentially help us to optimise the settlement of batches of securities transactions.

Another important aspect of the quantum computing work we are doing relates to risk mitigation, and how to behave responsibly in finding new uses for the technology. It is almost certain that quantum computing will bring new risks in future and our top priority will remain ensuring safety for our clients and partners, and for our business. We are therefore also exploring the role of ‘quantum resistant’ security systems and look forward to collaborating across the industry on best practices and standards.

It’s about running certain complex computer processes faster and more efficiently than on existing classical computers

‘Quantum computing’ describes the technological process where a computer takes advantage of some specific properties of subatomic particles in order to solve certain types of problems. It is able to do this faster and more efficiently than is possible on the sort of classical computers that you find in homes, offices and on trading floors at present.

It is governed by laws of quantum physics

Quantum computing exploits the rules of ‘quantum mechanics’, which is a theory that describes nature at the small scale of atoms and subatomic particles. Quantum physical laws are very different to both the classical physical laws that describe nature at the ordinary scale we normally perceive with the human eye and the relatively simple logic of classical computers. In quantum mechanics, atoms and subatomic particles can behave in an unpredictable manner that simply can’t be modelled using the deterministic rules of classical physics.

It allows us to derive greater value from our data

Quantum computing will allow us to address certain problems that are too complicated to be solved using today’s most advanced supercomputers. This could include, for example, evaluating more candidate answers to a problem in order to identity a better answer.

Quantum computers are made up of units of information called ‘qubits’

Quantum computing is all about leveraging a fundamental unit of information called a ‘qubit’. This is similar to a classical computer memory ‘bit’ that can be in a state that represents either 0 or 1, but the crucial difference is that a qubit can be in a ‘superposition’ of both states simultaneously and thereby represent both 0 and 1 simultaneously.

In quantum computing, many qubits can be linked together

A quantum computer includes a circuit comprising qubits that are linked together. And because each qubit can represent both 0 and 1 simultaneously, a collection of qubits that are tightly correlated (via a phenomenon we call ‘entanglement’) can contain exponentially more information than the same number of classical bits.

Quantum computing increases our processing power

When you perform operations on your data using quantum computing, you are typically performing calculations on many variations of each data item simultaneously. This is known as ‘quantum parallelism’ and is hugely valuable as it permits an exponential increase in processing power.

What the experts think

We are some way off from being able to leverage the full potential of quantum computing – estimates as to when we will be able to use it to solve real-world problems in financial services range between five and 10 years. Huge investments of time and resources are being made by organisations to develop quantum computing technology and bring forward the date when we will be able to use the technology in our day-to-day lives.

How we can use quantum computing now

The most advanced general-purpose quantum computers available today typically have only a few tens of qubits. And those qubits are usable for only very short periods of time (due to a phenomenon known as ‘quantum decoherence’) and are subject to high error rates. These restrictions impose limits on how much we can do with quantum computers today. We can use them to prove principles and run proofs-of-concept, but we can’t actually put them to work with real-world financial services problems yet.

What we still need to learn

To put quantum computing into action at scale, we need hardware with many more qubits – ideally millions of qubits. And they will also need to be resilient, which will require the introduction of effective error correction. However, it will probably be a further five to 10 years before such quantum computers can be built. In the meantime, organisations such as Barclays are building skills and knowledge to advance their understanding of how to use the technology, as well as collaborating with partners to explore potential business applications.

Safe for now, but a significant risk in future

Quantum computing is a new technology with the potential to bring new threats to business and society at large. There is a strong possibility that at some point in the future, bad actors such as cybercriminals could obtain access to quantum computers that are sufficiently powerful to crack classical cryptography. This means that unless business and society put mitigations in place, the security systems that exist today to protect private and confidential information could become redundant.

How do we know?

We can be certain that this risk exists because quantum algorithms have been invented which could, in theory, crack classical cryptography by efficiently finding prime factors. An example is ‘Shor’s algorithm’, which was invented over 25 years ago. However, it is estimated that it will be another 10 to 15 years until a sufficiently powerful quantum computer, with a few million qubits, may become available.

Actions being taken now

Businesses like Barclays are taking steps now to explore the potential risks associated with quantum computing. This includes gaining an understanding of the timeline and potential mitigations that will be required at some point in future to protect our customers and our systems from those risks. The solutions will involve upgrading the existing cryptography that secures systems to new ‘quantum-resistant cryptography’. We are monitoring progress in both quantum computing hardware and quantum-resistant algorithms, and we look forward to cross-industry collaboration on standards for quantum-resistant cryptography.

When could risks emerge?

There is a range of opinion amongst researchers and other experts as to when this risk could emerge. Estimates range from five to 30 years away, with some leading companies researching quantum computing estimating 10 to 15 years. The National Institute of Standards and Technology (NIST) in the US has suggested it may be a decade until a bad actor could obtain the capability to crack cryptography using a quantum computer.

Bad actors are developing threats now

In anticipation of the possibility that security systems could be hacked into using quantum computing in future, bad actors and cybercriminals may be hoarding business’ and individuals’ encrypted information. They can’t decrypt and access that data now but, once they are able to leverage quantum computing’s ability to crack cryptography in future, they may then be able to unlock and exploit the private information. One of the reasons that businesses like Barclays are monitoring progress in quantum computing is so we can be aware of timelines for such threats.

Why the industry is so interested in quantum computing

Quantum computing has the potential to transform many aspects of financial services. Its processing power could eventually address problems that are too complicated to be solved using today’s most advanced supercomputers. It is difficult to predict how this could enhance business processes and business workforces in future, but we expect it will be deployed initially to optimise existing processes. For example, to improve the efficiency of financial market analysis, risk management, collateral optimisation, and transaction settlement.

Collaboration within the industry

Some banks, such as Barclays, are actively exploring quantum computing. This includes gaining an appreciation of the potential, understanding the technology, identifying candidate business use cases, conducting experiments, and publishing research results. There is appetite within the financial services industry to collaborate on quantum computing because all parties stand to benefit from pooling knowledge that grows our collective understanding of this emerging technology.