A Pragmatic Approach to Adiabatic Quantum Computation

The quantum world is innately parallel. Quantum objects may exist in many places at the same time and in general have a superposition of attributes that would be mutually exclusive for an object on the everyday classical scale. In 1982 Richard Feynman suggested that one might attempt to use this parallelism to speed up computation and in 1982 Peter Shor discovered an algorithm that could, theoretically, make use of it in a calculation.

Since these early theoretical works, there has been a dramatic effort in the theory of quantum computation while at the same time trying to find a physical system where these ideas could be realized. Taking the queue from the success of digital, gate-based, classical computation, much of this effort has focused on gate based digital quantum computation. There is an alternative, however, which harnesses our understanding of physical process rather more directly.

Nature is rather good at solving problems such as finding the most efficient way to arrange a collection of atoms into a crystal. Nature achieves this by gradually reducing the temperature of a system so that it can eventually settle to its lowest energy state - a process known as thermal annealing. This is used in a range of classical optimization algorithms.

A quantum version of this, originally known as quantum annealing - now known as adiabatic quantum computation - may ultimately prove to be more effective for quantum computation than the gate based model. Indeed, a Canadian company, D-wave Systems, has attempted to make just such a computer with some promising initial results. Interpreting such attempts is difficult, however, since the failure mode of an adiabatic quantum computation is a classical thermal anneal.

This project aims to develop a systematic way to test whether an adiabatic quantum computation has taken place using a pragmatic, physics based approach. In doing so, new insights into how to optimize the performance of such a system will result.

Adiabatic quantum computation sits at the focus of several different areas: i. the fundamental physics of quantum systems out of equilibrium, ii. Quantum information and computer science, and iii. A broad base of potential users in areas as diverse as science finance, business and statistical consultancy, medicine and logistics.

The impacts in science and quantum information will be immediate. Out-of-equilibrium quantum systems have been identified as a grand challenge both in the US and the UK. Adiabatic quantum computation is a good example of interest to a broad range of physicists. The related classical annealing algorithms have benefitted greatly from physical understanding of spin glasses; insights about how to find the minimum energy state have led directly to new algorithms. Adiabatic quantum computation is ripe for a similar type of analysis. This approach will complement the progress made by quantum information theorists in understanding quantum computation.

Many areas of key economic and societal importance to the UK rely upon complex optimization and efficient computing resources to do it. The proposed work will seed a firm skills base for the development of future applications.

The links with Lockhead-Martin/USC will be of particular benefit in enhancing impact. It provides a new and uniquely controllable forum to investigate out of equilibrium quantum systems; an important task of the PDRA will be to learn to programme their Dwave machine. Indeed, engagement with Lockhead-Martin/USC will provide a clear link between theory, practice and application.

UCL and the LCN provide an ideal location for this project. The director of the LCN, Prof Gabriel Aeppli, was a PI on the early experimental demonstration of the principle of adiabatic quantum computation. In addition UCL and the LCN host world-leading theorists working in quantum information and so provides opportunity for cross-fertilization with the pragmatic, physics-based approach espoused here.
Several other groups in UCL are interested in adiabatic quantum computation and the university is slowly beginning to build a critical mass in this area. In particular, Dr Paul Warburton of Electrical Engineering and Physics, and Dr Simone Severini of Computer Science and Physics will shortly appoint a post-doc under EPSRC Global Engagement funding to explore applications of adiabatic quantum computation. This post-doc will complement the more fundamental aims of this programme. In March 2013, UCL will host the 2nd International Workshop on Adiabatic Quantum Computation.

The PI is engaged with broader efforts in the study of out-of-equilibrium quantum systems both through his core research supported by and EPSRC Leadership Fellowship and through his involvement in the TOPNES programme grant. Collaborative links with researchers experienced in the area of spin glasses - in particular with Prof Juan P Garrahan of Nottingham - are being forged to aid and benefit from this programme.

Propagation of results will be through the usual avenues of publication in high impact journals and conference presentation. With an eye to future application, the PI and LCN are forming a network of interested potential users in science, medicine, finance, business and statistical consulting. The business development officer in the LCN will play an active role in maintaining and developing these links. The general public will be kept informed of progress by accessible posts on the LCN website, occasional public lectures and interaction with popular science journals. The PI has extensive experience of public engagement with lectures at various Cafés Scientifique throughout the UK, at the Royal Institution as well as founding and running a Café Scientifique in St Andrews. The post-doc will be encouraged to engage with the public through forums such as UCL-based cosy science and other UCL outreach programmes.