Skyrmion-based Electronics

Project description

Future integrated systems will comprise considerably heterogeneous/hybrid technologies both for computing and (on-chip) data storage. One of the emerging technologies that can complement CMOS technologies is spintronics. The distinct characteristic of spintronics is that information is retained and processed by utilizing the spin rather than the charge of the electrons [1].
A potential spin-based technology, magnetic skyrmions, has attracted extensive interest due to their unique physical characteristics. Magnetic skyrmions are nanoscale topologically stable vortex-like spin windings that are robust due to their topology and are considered promising candidates for information carriers in future computing systems. Their small size (nanoscale), robustness against material defects and low electrical currents required to manipulate them can lead to next generation ultra-dense, robust and low-power spintronic devices [2]. Skyrmions have been demonstrated experimentally at room temperature for the first time in technologically relevant multilayers [3], which opens the way for their use in novel applications [4]. Our team has also made significant strides in devising interconnect structures and designing neuromorphic circuits with skyrmions. We are presently developing novel devices, models, and circuits towards the vision of integrating skyrmion-based computing with mainstream CMOS technologies facilitated by the UKRI grant "Skyrmionics for Neuromorphic Technologies".

The main goal of this PhD project is to develop fundamental components for skyrmion-based electronics and related compact models for circuit simulation. The project builds on recent achievements of the group [5] in this research area where novel skyrmionic devices for neural networks [6] and interconnects [7] were demonstrated.

Consequently, the project aims to exploit these milestones to demonstrate one or more of the following: i) enhanced components that utilize the non-volatile nature of skyrmions (and other related topological quasiparticles) to highly improve the energy-efficiency of computing and/or data transfer, ii) circuit level models for these components that are highly scalable and require reasonable simulation times, and iii) use these components for emerging unconventional architectures beyond the Von Neumann paradigm (e.g., in-memory, probabilistic and neuromorphic computing).

Research aiming to these objectives is expected to lead to high impact publications in both international competitive conferences and esteemed journals.