Photonic Integrated Circuits (PICs) are emerging as a transformative platform for next-generation artificial intelligence (AI) hardware, enabling ultrafast, energy-efficient optical data processing. However, as PICs scale to the densities required for AI acceleration, thermal effects become a critical bottleneck, limiting performance, reliability and efficiency.
This challenge presents a global sustainability problem since data infrastructure is projected to consume over 20% of worldwide energy within a decade. Tackling this demands a step change in how temperature is measured and controlled directly on-chip. This PhD ...