Dr Rostislav Mikhaylovskiy, Lancaster University

Cutting the heat from data centres

The challenge

Data centres are expanding rapidly to support our soaring use of technology – specifically wireless devices and AI applications, all of which require a huge volume of cloud storage. But how can we expand data centre capacity while tackling the sustainability challenges surrounding their energy, water and resource use? Dr Rotislav Mikhaylovskiy is developing a technology that allows for this expansion, while cutting data centres’ energy use – and in turn the heat they generate. 

Data centres require huge amounts of energy to function, in large part because they produce so much heat that must be cooled to enable them to function effectively. With the expansion of AI, it’s predicted that there will be a 160% increase in data centre power demand globally by 2030. 

The innovation

Dr Mikhaylovskiy is developing a new way of storing data, which uses extremely short bursts of terahertz radiation – light pulses a thousand times faster than today’s technology. These pulses flip the direction of small magnets that store data. Because these pulses match the magnets’ energy, they can switch them without creating heat. The goal is to make data storage much more efficient, limit the amount of heat that this process produces, and therefore reduce the huge volume of energy that data centres demand to keep servers cool.

With the Green Future Fellowship funding, Dr Mikhaylovskiy is recruiting and training a research team to develop, prove and scale this technology for real-world impact.

The innovator

Dr Mikhaylovskiy completed his PhD at the University of Exeter, where his work focused on how light can influence and control magnetic systems. Following this, he moved to Radboud University in the Netherlands, where he deepened this expertise. Today, he leads a research group at Lancaster University, which is dedicated to ultrafast magnetic switching and the interaction of magnets with short laser pulses.

• Web page

The fellowship has the potential to disrupt a major issue facing AI today, helping to launch the UK into first position in the race to be the great AI leader.