Photonic Quantum Computation

Image credit: Alessandro Fedrizzi

Why photons? Over the past decade, a wide variety of physical architectures have been investigated for their suitability for quantum technologies, including trapped atoms and ions in ultra-high vacuum, superconducting circuitry and photonic systems [1]. The critical requirements for any architecture are: quanta that can be well-isolated from the environment so that information encoded in them is not degraded by noise; the ability to control and store these quanta and the capability to interact them when required, via entangling-gates. (Entanglement produces correlations between systems that cannot be replicated using classical physics; it is a powerful and uniquely quantum phenomenon, famously described by Einstein as "spooky action at distance".) Photons have been very successful in meeting these criteria, and indeed for some applications—such as communication, remote sensing and networking of quantum information—photons are not only the natural, but the only choice. They are move at the speed of light, and retain quantum correlations extremely well as long as they are not absorbed. Photonic entangling-gates have been continually improved since their first demonstration in 2003 [2] and now routinely achieve high fidelities [3]. Photonic systems have been used in demonstration systems in computation and metrology, and are now the basis of commercial systems in quantum cryptography, with companies in Australia, France, Switzerland and the US [4].

Image credit: Stewart Gould

The challenge. Despite their success, photonic quantum technologies currently face a significant roadblock to widespread application: the circuits are severely limited in complexity. One measure of circuit complexity is the product of its breadth (number of photons), and its depth (number of interactions between the photons). To date, the best results for photonic quantum technology, achieved in separate experiments, are a breadth of 6 photons and a depth of 2 entangling interactions. For many applications, this does not provide enough complexity to gain an advantage over existing technologies. This size limitation is particularly frustrating because an order-of-magnitude increase in complexity is predicted to deliver the capability to perform many useful tasks. Since the information capacity in quantum technologies scales exponentially, this yields not a 10-fold, but rather a 2^10-fold increase in capacity. Even modest numbers of photons and entangling interactions, on the order of tens to hundreds, would be game-changing both scientifically and technologically.

Increasing circuit complexity will require solutions to the following inherent difficulties:

1) It is difficult to store photons, since they interact weakly with other particles and move at the speed of light. This limits circuit breadth, since many protocols require holding information in one part of the circuit while waiting for information to be processed in parallel.

2) It is difficult to efficiently produce and detect single photons. The current best photon sources are spontaneous, i.e. the photons are produced at random times with probability, p<1. This quickly limits circuit breadth, since the probability of producing 1 photon per mode decreases exponentially (for N input modes it is p^N << 1).

3) Current photonic entangling gates are inherently random—with success rates varying between 9% and 25% [2]—which means they cannot be scaled. This severely limits circuit depth: if one gate works 1/4 of the time, then two gates in sequence will only work 1/16 of the time, and so on. In principle, the gates can be made scalable at the cost of significant architectural overhead: in the original proposal from 2003, each logical gate required 100,000 physical gates [5]; recent proposals have reduced this to 10 to 100 physical gates. Given the severe technical difficulties, none of these proposals has yet been demonstrated in practice.

Image credit: Stewart Gould

Why gas-phase? We will address these difficulties by taking a different approach to quantum photonics. In the current paradigm of quantum photonics, the only nonlinear elements are the detectors, since conventional nonlinear optical materials are orders-of-magnitude too weak to allow direct photon-photon interaction [6]. Even cavity-QED systems do not provide a strong enough interaction, despite their strong photon-atom coupling, and of course they are too fragile, expensive and bulky to consider as any kind of scalable solution. We will pursue a hybrid approach, exploiting the strong single- and two-photon absorption possible in the gas-phase of rubidium atoms, together with integrated-photonics, to achieve strong interactions between photons and atoms, and use these interactions to achieve efficient quantum memories, efficient photon detectors, and reliable entangling gates.

[1] M.A. Nielsen and I.L. Chuang, Quantum computation and quantum information (Cambridge University Press, UK, 2000).

[2] J.L. O'Brien, G.J. Pryde, A.G. White, T.C. Ralph, and D. Branning, Nature 426, 264 (2003).

[3] B.P. Lanyon, M. Barbieri, M.P. Almeida, T. Jennewein, T.C. Ralph, K.J. Resch, G.J. Pryde, J.L. O'Brien, A. Gilchrist and A.G. White, Nature Physics 5, 134 (2008).


[5] E. Knill, R. Laflamme, and G.J. Milburn, Nature 409, 46 (2001).

[6] G.J. Milburn, Physical Review Letters 62, 2124 (1989).