Professor Will Branford Inaugural Lecture

Join Professor Will Branford, Professor in Physics, for his Imperial Inaugural online or in person.

There is no need to register to attend so please be sure to use the add to calendar button.

We look forward to seeing you on Wednesday 26 February!

Summary

The algorithms that power modern Artificial Intelligence (AI) emerged from basic physics research simulating complex systems (Physics Nobel Prize 2024). The incredible success of AI comes with an astonishing energy cost. Data centres already consume 3% of global energy output, increasing quickly. Leading companies are solving this problem using traditional computers and purchasing power plants next to datacentres.

This is not sustainable, at current growth, and the planet will not be able to produce enough energy for AI in 10 years’ time. We know there is a path forward: the human brain consumes 12W to do tasks AI requires billions of watts to perform.

Will Branford is a Professor in Physics at Imperial College London where, inspired by biological processes, his research focuses on neuromorphic computers, for which computing is not the result of simple mathematical operations repeated at extremely high speed but instead a few operations on a complex physical system. In his inaugural lecture he will explore the potential for this brain inspired systems, as well as magnonics, where the information is carried in spin-waves.

Biography

Will Branford is a Professor in Physics and the Head of Community for Matter. He is a Co-director of the School of Human and Artificial Intelligence. He is part of the management group of the London Centre for Nanotechnology and active in the nanotechnology effort at Imperial.

His main research focus is the interplay between structure, geometry, magnetic and electronic properties of materials. His background is in solid state synthesis of magnetic materials, where the structuring was at the atomic scale. His primary focus now is in magnetic textures at the nanoscale, where nanofabrication can impose an additional lengthscale of structure and geometry. Artificial Spin Ice is an example of a magnetic metamaterial whose properties derive from the properties of the nano-units in the array, rather than the material it is made from.

A strong motivation for the research is exploring the potential of new methods of computation, such as neuromorphic computation, which is inspired by the brain, and magnonics, where the information is carried in spin-waves.

Getting here