Cooling in data centers, which is highly energy-intensive, accounts for 40% of their total energy consumption. This contributes to 2% of global energy use and 2.5% of greenhouse gas emissions, a situation that is expected to worsen with the rapid expansion of AI. Our goal is to enhance energy efficiency by improving the thermal conductivity of chips.

My co-founders and I initially became heavy AI users due to being non-native speakers, and we were fascinated by its power and its potential to improve our quality of life. However, as we delved deeper into the energy consumption of AI, we were shocked by the enormous energy demands and became deeply concerned about the sustainability of our planet. Therefore, as materials scientists, we aim to use our expertise to make AI more sustainable.