Periodic Reporting for period 2 - VLS-QPP (Very-Large-Scale Quantum Photonic Processing)
Reporting period: 2019-04-01 to 2020-03-31
The rules of quantum mechanics promise information processing technologies that are inherently more powerful than their classical counterparts: examples include quantum computing, unconditionally secure communication, and quantum-enhanced precision sensing. After several decades of intensive theoretical and experimental efforts, the field of quantum information processing is on the cusp of a critical moment: over the next decade, quantum computers and special-purpose quantum information processors will be capable of solving problems that classical computers cannot. The possibility of powerful new quantum information processors has ignited major investment around the globe by governments (viz. the EC’s €1B Quantum Technologies Flagship, China’s Quantum Satellite) and industry (Google, Microsoft, DWave, IBM). Photons play a central role in quantum computing and quantum networks due to their low noise properties, excellent modal control and long distance propagation. Photonic integrated circuit technology has enabled orders of magnitude improvements in component density, propagation loss, and phase stability. These advances have made possible proof-of-principle demonstrations of central quantum protocols, such as (compiled) factorization and quantum simulation of energy levels of small molecules.
Advancing the field to computationally hard problems requires a new generation of ‘quantum photonic processor’ that efficiently integrates nonclassical light generation, high-fidelity mode transformations and nonlinear photon-photon interactions. As quantum photonic technology advances new applications areas must be discovered where near-term quantum processors will likely have impact. This symbiotic development of hardware and algorithms is a central tenet of this program. ‘Very-Large-Scale Quantum Photonic Processors’ develops the next generation of quantum optical technology using the platform of silicon photonics. Silicon photonics leverages large-scale silicon manufacturing and CMOS technology to develop micron-scale photonic structures at an unprecedented component density and scale.
Applying silicon photonic technology to the quantum regime, and at a scale where classical computers can no longer keep pace, requires breakthroughs in quantum photonic engineering. Very-Large-Scale Quantum Photonic Processors addresses this challenge on two fronts: First, by leveraging state-of-the-art low-loss silicon photonics and advances in large-scale packaging and control, quantum photonic processors will be scaled-up to tackle problems at the limit of what is classically simulable. Second, by integrating solid-state quantum emitters with silicon photonics a scalable path is put-forth to enable the deterministic generation of photons and strong photon-photon interactions. In concert, a new arsenal of quantum algorithms will be developed specifically for implementation on this new generation of quantum photonic processors, targeting the key application areas of machine learning and quantum simulation.
Advancing the field to computationally hard problems requires a new generation of ‘quantum photonic processor’ that efficiently integrates nonclassical light generation, high-fidelity mode transformations and nonlinear photon-photon interactions. As quantum photonic technology advances new applications areas must be discovered where near-term quantum processors will likely have impact. This symbiotic development of hardware and algorithms is a central tenet of this program. ‘Very-Large-Scale Quantum Photonic Processors’ develops the next generation of quantum optical technology using the platform of silicon photonics. Silicon photonics leverages large-scale silicon manufacturing and CMOS technology to develop micron-scale photonic structures at an unprecedented component density and scale.
Applying silicon photonic technology to the quantum regime, and at a scale where classical computers can no longer keep pace, requires breakthroughs in quantum photonic engineering. Very-Large-Scale Quantum Photonic Processors addresses this challenge on two fronts: First, by leveraging state-of-the-art low-loss silicon photonics and advances in large-scale packaging and control, quantum photonic processors will be scaled-up to tackle problems at the limit of what is classically simulable. Second, by integrating solid-state quantum emitters with silicon photonics a scalable path is put-forth to enable the deterministic generation of photons and strong photon-photon interactions. In concert, a new arsenal of quantum algorithms will be developed specifically for implementation on this new generation of quantum photonic processors, targeting the key application areas of machine learning and quantum simulation.
VLS-QPP simultaneously develops new quantum photonic hardware alongside a suite of quantum algorithms and protocols specifically designed for next generation quantum photonic processors. Consequently, the primary results straddle both experiment and theory. On the hardware side, a large-scale silicon-based quantum photonic processor was designed, developed and tested, and new quantum machine learning algorithms were demonstrated. Alongside systems level engineering, critical components such as modulators, photon sources and single-photon nonlinearities were also developed. On the theory side a new neural network inspired architecture for near-term quantum photonic computing was proposed, and a new quantum machine learning algorithm 'variational quantum unsampling' was proposed and demonstrated on a quantum photonic processor. Results were disseminated via 29 invited talks across 3 different continents; 8 publications including 2 first author, 1 last author and two Nature family; and the development of a hands-on quantum computing course.
This program develops quantum processors at an unprecedented scale, and lays the foundation for scalable photonic quantum computing. This new class of quantum processor, with new computational capabilities in quantum simulation and machine learning, will likely have a transformational impact on the fields of computing and machine learning by enabling a scalable and robust hardware platform that can implement an entirely new family of near-term quantum algorithms. In particular, there are at least three key fields, with strong potential for societal impact in the EU and beyond, that stand to benefit the breakthroughs in this work: (1) Medical science through the design of new molecules; (2) Data security through the development and implementation of new and more robust quantum repeater architectures, (3) Machine vision by directly performing inference on optical signals from LIDAR systems. These advances will in turn engender new market opportunities, and both create and grow companies in the ever expanding quantum industry sector and beyond.