Emulation and optimisation of circuits for quantum simulation

Lead Research Organisation: University College London
Department Name: London Centre for Nanotechnology

Abstract

Quantum computers are expected to provide an exponential speedup over traditional computers in performing certain tasks. As experimentalists begin to build prototype devices, it would be useful to be able to benchmark and verify them against simulations run on classical computers, but the purported exponential advantage of quantum computation presents an obvious practical barrier to doing so. On the other hand, it is known that a restricted class of quantum circuits, known as stabiliser circuits, can be efficiently simulated using only classical computational resources. It is expected that the classical simulation cost for more general quantum circuits will be in some sense exponential in the amount of non-stabiliser operations required. Related is the magic state model of quantum computation, where non-stabiliser resource states known as magic states are "injected" into a circuit in order to implement non-stabiliser operations. In many schemes for fault-tolerant quantum computation, the requirement for highly pure magic states produces a large experimental overhead whenever non-stabiliser operations are needed. It is therefore desirable to rigorously quantify the non-stabiliser resource involved in a given circuit. Earl Campbell and Mark Howard at the University of Sheffield previously developed a resource theory of non-stabiliser states based on a measure called robustness of magic. For those operations that can be implemented using magic state injection, the theory immediately yields a means to quantify their classical simulation overhead. Only certain operations can be implemented in this way, however, and it is less clear how best to quantify the non-stabiliser resource involved in more general circuits. The current focus of my project is to extend the framework to deal with general quantum operations. Longer term, we hope to use insights from this work to develop improved algorithms for simulating general quantum circuits, and to optimise circuit design to consume less non-stabiliser resource. We believe that the scenarios where these techniques might be most usefully applied include noisy quantum devices and certain quantum chemistry simulations.

Planned Impact

Quantum technologies promise a transformation of the fields of measurement, communication and information processing. They present a particular opportunity since they are disruptive technologies: not only do they offer a chance for rapid growth but they also allow lesser participants in a field (such as the UK in IT) to become major players through appropriate risk-taking and manpower development. Students graduating from the InQuBATE Skills Hub will have the right mindset to work in the industries where quantum technologies will be applied, and help to break down the traditional barriers between those sectors to make this transformation happen. They will have all the necessary technical and transferable skills, plus a network of contacts with our partners, their fellow cohort members and the academic supervisors.

Our commercial partners are keen to help our students realise their potential and achieve the impact we expect of them, through the training they offer and their contributions to the centre's research. They include companies who have already developed quantum technologies to products in quantum communication (Toshiba) and optimization (D-Wave), large corporates who are investing in quantum technology because they see its potential to transform their businesses in aerospace, defence, instrumentation and internet services (Lockheed Martin, Google,) and government agencies with key national responsibilities (NPL). We want to see the best communication of our students' research, so our students will benefit from the existing training programme set up with a leading scientific publisher (Nature Publishing Group); we also want to see more of the future companies that lead this field based the UK, so we have partnered with venture capital group DFJ Esprit to judge and mentor the acceleration of our students' innovations toward the market.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/P510270/1 01/04/2016 31/08/2022
1781317 Studentship EP/P510270/1 01/10/2016 28/02/2021 James Seddon
 
Description It is widely anticipated that large scale quantum computers will be able to perform certain tasks that are intractable on traditional (classical) computers. On the other hand, it is known that there is a large class of quantum circuits, called stabiliser circuits, that can be efficiently classically simulated. Between these two extremes, there is an intermediate regime of near-stabiliser circuits where classical simulation is likely to be inefficient, but still at least tractable given access to enough conventional computing power for a reasonable timescale. The simulation cost is expected to scale close to exponentially with the quantity of non-stabiliser resource, known as magic, needed by the circuit. Two of the aims of this project were (i) to rigorously quantify this resource for general noisy quantum operations on qubits and (ii) to develop improved classical simulation algorithms for near-term noisy intermediate-scale quantum (NISQ) devices. To this end, we have:

(1) defined two families of well-behaved measures of non-stabiliser resource for operations, one family based on different types of channel decomposition, and one based on capacity to generate resource;

(2) proposed two families of classical simulation algorithm for general quantum operations on qubits, where the performance is quantified by our resource measures. We improve on previous related methods, demonstrating a significant reduction in runtime for certain types of quantum circuit;

(3) in collaboration with academic partners, developed improved variants of two classical simulation algorithms for stabiliser circuits with magic state inputs, based on new measures of magic for states, again demonstrating significantly reduced runtime compared to previous methods;

(4) conducted numerical investigations to identify parameter regimes in which particular simulators show an advantage over another.
Exploitation Route A number of groups in academia and industry are working towards building intermediate-scale quantum computers. In the near-term these devices are likely to suffer from noise. Our findings could be used to optimise the efficiency of classical simulators used to benchmark and verify these noisy intermediate-scale quantum devices. Our techniques could also be of interest to quantum technology startups interested in classical simulation of quantum devices.
Sectors Digital/Communication/Information Technologies (including Software),Other

URL https://doi.org/10.1103/PRXQuantum.2.010345
 
Description Quantifying quantum speedups: improved classical simulation from tighter magic monotones 
Organisation Nanyang Technological University
Department School of Physics and Mathematical Sciences
Country Singapore 
Sector Academic/University 
PI Contribution In collaboration with my supervisor I contributed my expertise in classical simulation and the resource theory of magic. I developed new variants of classical simulation algorithms, drafted sections of the manuscript on this topic, critically revised the manuscript, and prepared and presented talks on our work at several academic conferences.
Collaborator Contribution Our collaborators provided expertise on classical simulation of quantum of circuits, general resource theories and convex optimisation. All collaborators were involving in the drafting and critical revising of the manuscript.
Impact The collaboration has resulted in an article accepted for publication in PRX Quantum, "Quantifying quantum speedups: improved classical simulation from tighter magic monotones."
Start Year 2018
 
Description Quantifying quantum speedups: improved classical simulation from tighter magic monotones 
Organisation Perimeter Institute for Theoretical Physics
Country Canada 
Sector Academic/University 
PI Contribution In collaboration with my supervisor I contributed my expertise in classical simulation and the resource theory of magic. I developed new variants of classical simulation algorithms, drafted sections of the manuscript on this topic, critically revised the manuscript, and prepared and presented talks on our work at several academic conferences.
Collaborator Contribution Our collaborators provided expertise on classical simulation of quantum of circuits, general resource theories and convex optimisation. All collaborators were involving in the drafting and critical revising of the manuscript.
Impact The collaboration has resulted in an article accepted for publication in PRX Quantum, "Quantifying quantum speedups: improved classical simulation from tighter magic monotones."
Start Year 2018
 
Description Quantifying quantum speedups: improved classical simulation from tighter magic monotones 
Organisation University of Sheffield
Department Department of Physics and Astronomy
Country United Kingdom 
Sector Academic/University 
PI Contribution In collaboration with my supervisor I contributed my expertise in classical simulation and the resource theory of magic. I developed new variants of classical simulation algorithms, drafted sections of the manuscript on this topic, critically revised the manuscript, and prepared and presented talks on our work at several academic conferences.
Collaborator Contribution Our collaborators provided expertise on classical simulation of quantum of circuits, general resource theories and convex optimisation. All collaborators were involving in the drafting and critical revising of the manuscript.
Impact The collaboration has resulted in an article accepted for publication in PRX Quantum, "Quantifying quantum speedups: improved classical simulation from tighter magic monotones."
Start Year 2018
 
Description Quantifying quantum speedups: improved classical simulation from tighter magic monotones 
Organisation University of Sydney
Country Australia 
Sector Academic/University 
PI Contribution In collaboration with my supervisor I contributed my expertise in classical simulation and the resource theory of magic. I developed new variants of classical simulation algorithms, drafted sections of the manuscript on this topic, critically revised the manuscript, and prepared and presented talks on our work at several academic conferences.
Collaborator Contribution Our collaborators provided expertise on classical simulation of quantum of circuits, general resource theories and convex optimisation. All collaborators were involving in the drafting and critical revising of the manuscript.
Impact The collaboration has resulted in an article accepted for publication in PRX Quantum, "Quantifying quantum speedups: improved classical simulation from tighter magic monotones."
Start Year 2018
 
Description Quantifying quantum speedups: improved classical simulation from tighter magic monotones 
Organisation University of Waterloo
Country Canada 
Sector Academic/University 
PI Contribution In collaboration with my supervisor I contributed my expertise in classical simulation and the resource theory of magic. I developed new variants of classical simulation algorithms, drafted sections of the manuscript on this topic, critically revised the manuscript, and prepared and presented talks on our work at several academic conferences.
Collaborator Contribution Our collaborators provided expertise on classical simulation of quantum of circuits, general resource theories and convex optimisation. All collaborators were involving in the drafting and critical revising of the manuscript.
Impact The collaboration has resulted in an article accepted for publication in PRX Quantum, "Quantifying quantum speedups: improved classical simulation from tighter magic monotones."
Start Year 2018