Quantum Optical Routes to Quantum Supremacy


May 03, 2017

Speakers: Sir Peter Knight


Durham University

About this event

This seminar will be held in PH8 @ 12:00 - 13:00
Quantum states of many particles offer the opportunity to study the rich physics of large-scale quantum correlations. Such multi-partite correlations are believed to underlie a range of physical phenomena such as structure and transport in many-body quantum systems and promises performance that dramatically exceeds any classical processor. Building machines able to achieve this so-called Quantum Supremacy has obvious motivation from a technological standpoint [1]. Systems in which all particles and their interactions can be controlled to some degree can deliver functionality for sensing, imaging, communications, simulation and computation that goes well beyond what is possible with classical systems possessing a similar set of resources.
One approach is to build large quantum states of light, constructing quantum correlated light beams, with a system comprising many photons in which each input and their interactions are controllable. The rationale for using light is three-fold. First, photons are one of a very few systems that exhibit palpable quantum phenomena under ambient conditions, and therefore the engineering required to produce and control particles in identical pure quantum states is minimized. Second, the number of bosons required to enter a regime that cannot be simulated is relatively small [2]: several tens to a hundred photons in a similar number of modes. Third, a light beam becomes correlated with another at the quantum level by means of a fundamental quantum interaction – exchange – manifested as quantum interference. Indistinguishability and the exchange interaction play a central role in the emergence of entanglement and other stronger-than- classical correlations in the interference of many quantum particles. Photons are quite unique in that, having no charge, they are purely influenced by their particle statistics and provide an ideal opportunity to study the effects of purity and distinguishability on quantum interference, isolated from any contributions from external forces. One important development in this direction was the formulation of the Boson Sampling problem [3]. As originally envisioned – sampling the output distribution of identical Fock states scattered through a linear optical network [2] – Boson Sampling is unique: it implements a classically hard problem but with massively reduced resource requirement compared to more traditional quantum technologies. I will explain how simple quantum networks implement quantum walks and Boson Sampling.
This was supported by the EPSRC Programme BLOQS (Building Large Quantum States out of Light), a collaboration between Oxford, Southampton and Imperial College.
References [1] P Kok, Contemp Phys 1178472 (2016); arXiv:1603.05036 and references therein.
[2] S Scheel, K Nemoto, W J Munro and P L Knight, Phys Rev A68, 032310 (2003) for an early demonstration of quantum supremacy in a linear quantum optical network, showing how output distributions are described by permanents, hard to evaluate by classical means.  [3] Aaronson & Arkhipov Proc. ACM sym. 333-342 (2011).

Back to full list of seminars.