معرفی
Dr. Matt Ellis is a Lecturer in Machine Learning at the University of Sheffield, affiliated with the Machine Learning Group within the Department of Computer Science. He is also the Director of UG Admissions and Admissions Tutor for the School.
Education
- PhD in Physics (2015), University of York
- MPhys in Theoretical Physics (2011), University of York
Research Interests focus on neuromorphic computing and spintronics, particularly energy-efficient machine learning through physical reservoir computing and magnetic systems. His work explores using nanomagnetic devices as processing units, leveraging stochastic synapses, magnetic domain walls, and GPU-accelerated modeling. He investigates terahertz phonon pumping for minimally dissipative switching and emergent magnetization dynamics for computational capability.
Publications span 2014–2024, with recent articles on magnetic stochastic synapses, device-agnostic network optimization, and voltage-controlled superparamagnetic ensembles. His research integrates machine learning with physical computing to address challenges in edge computing, speech recognition, and chaotic time series prediction.
Grants
- Spintronic Reservoir Fusion (2025–2027), EPSRC, £563,599 (Principal Investigator)
- MARCH: Magnetic Architectures for Reservoir Computing Hardware (2021–2025), £936,815 (Co-Investigator)
- From Stochasticity to Functionality: Probabilistic Computation with Magnetic Nanowires (2019–2023), £755,424 (Co-Investigator)
Contact: m.o.ellis@sheffield.ac.uk
