Controlling quantum spin-1/2 networks for communication and computation
The aim of this seminar is to discuss quantum spin-1/2 networks.
Location: Room C/2.07, Queen's Buildings, School of Computer Science and Informatics.
Date: 7 February 2018 at 15:00
Abstract
The miniaturisation trend described by Moore’s law has led to nanoelectronic devices boosting the digital revolution for over five decades, from the millimetre scale to ten nanometres.
The trend slowed in 2009 and, as device features approach the size of atoms, scaling is going to further slow down and stop. As the feature size becomes smaller than 7nm, the performance of logic gates deteriorates due to device variability and devices can no longer robustly realise algorithms relying on the ideal operation of perfectly synchronous logic gates.
Harnessing the full range of quantum effects is one of the leading candidate technologies to deliver further increase of computing performance, including higher communication bandwidth, reduction of energy consumption and advanced sensing capabilities, which cannot be sustained by scaling of devices alone.
Networks of spin-1/2 particles form a generic prototype for such quantum devices. Information stored in spin states can propagate through these networks without any charge transport in a wave-like manner, dispersing and refocusing over time. The resulting traffic shows surprising non-classical behaviour, such as the presence of an anti-core, a node with minimal traffic instead of the congestion core in traditional networks.
Static energy landscape shaping is highly effective for controlling the flow of information through such networks. Optimal control algorithms find high-fidelity energy landscape controllers, similar to an optimising compiler realising a desired quantum routing or computation operation in a single step (instead of building the operation from quantum gates with error correction). Some of the resulting controllers show the astounding property of being optimally robust at optimal performance. However, robustness results also suggest performance and scalability limits, as classical behaviour emerges.
Biography
Dr. Frank C. Langbein received his mathematics degree from Stuttgart University, Germany, in 1998 and a PhD on 'Beautification of Reverse Engineered Geometric Models' from Cardiff University in 2003.
He is currently a senior lecturer at the School of Computer Science and Informatics where he is a member of the Visual Computing research group and leads the Quantum Technologies and Engineering (QuTeE) research priority area.
Dr. Langbein's research interests include modelling, simulation, control and machine learning applied to quantum technologies, geometric modelling, computer graphics, computer vision and healthcare problems.
