Perceval: A Software Platform for Discrete Variable Photonic Quantum Computing

Abstract

We introduce Perceval , an evolutive open-source software platform for simulating and interfacing with discrete-variable photonic quantum computers, and describe its main features and components. Its Python front-end allows photonic circuits to be composed from basic photonic building blocks like photon sources, beam-splitters, phase-shifters and detectors. A variety of computational back-ends are available and optimised for different use-cases. These use state-of-the-art simulation techniques covering both weak simulation, or sampling, and strong simulation. We give examples of Perceval in action by reproducing a variety of photonic experiments and simulating photonic implementations of a range of quantum algorithms, from Grover's and Shor's to examples of quantum machine learning. Perceval is intended to be a useful toolkit both for experimentalists wishing to easily model, design, simulate, or optimise a discrete-variable photonic experiment, and for theoreticians wishing to design algorithms and applications for discrete-variable photonic quantum computing platforms.

Strong Simulation of Linear Optical Processes

“…Using framework Perceval [18], we have implemented 6 the simulation of [13], using a generic m × m interferometer and its full output probability distribution to train a model solving differential equations. For an evolving configuration of the circuit, the algorithm has to iterate on all output states |t n m corresponding to input state |1, 1, ..., 1 n .…”

“…We compare in Figure 4 the speed of our algorithm to compute a single permanent in the worst possible situation (output is |1, ...., 1 ), with traditional permanent calculation algorithm. We selected the algorithm from Glynn [14] as [18] shows that for up to 15 modes, this algorithm is practically the most efficient. The graph shows that SLOS performs on par with the Glynn showing that no practical overhead is observed.…”

“…As a by-product, we derive an alternative algorithm for computing the permanent of a complex matrix that matches the state-of-the-art. We detail optimised implementations of our algorithms that can be found in the QuandeLibC library 1 (which is also integrated in the open-source software platform Perceval [18]). We benchmark the performance of our algorithms, comparing them to implementations of the Glynn and Ryser algorithms in the the Walrus library [17], as well as implementations in QuandeLibC of these algorithms, which appear to be more efficient, and of the Clifford and Clifford algorithms also in QuandeLibC.…”

Implementation of photon partial distinguishability in a quantum optical circuit simulation

“…Software focused on optical platforms is primarily centered on continuous variable models where observables of the electromagnetic field are calculated [13] and is therefore outside of our purpose. A recently released white paper [14] has announced the development of a software library Perceval which also considers a Fock description of photonic states. However, the available documentation of Perceval, including both the white paper and its github release, provides no technical information regarding many aspects of the Perceval implementation in general and the treatment of photon distinguishability in particular.…”

DisCoPy for the quantum computer scientist

“…An exciting direction of work in progress is to go away from the world of qubits altogether and implement the diagrams for linear optical quantum computing. DisCoPy's photonics module (still under development) will allow to build linear optical circuits and interface with Perceval [33] for their high-performance classical simulation. Implementing recent theoretical progress on a graphical calculus for linear optics [34], DisCoPy functors will allow to compile qubit circuits and ZX diagrams into linear optical circuits, one step on the photonic roadmap towards fault-tolerant quantum computing.…”