School of Engineering News
High-Efficiency Power Electronics for AI Data Centres and Net-Zero Grids
Data centres powering modern artificial intelligence workloads currently account for more than 2% of global electricity consumption, a figure projected to rise substantially over the next decade. In hyperscale facilities, around 30% of input electrical energy is lost as heat during power conversion and distribution.
To address the power conversion challenges outlined in the UK鈥檚 Clean Power 2030, Solid-State Transformers (SSTs) that combine wide-bandgap power semiconductors and high-frequency magnetics offer the power density, efficiency, and bidirectional power-flow control required for AI data centres, megawatt-scale electric vehicle charging, and renewable-rich distribution networks.