News -
How a Chinese Innovator is Aiming to Tame AI’s Power-Hunger | Falling Walls Science Summit, 6–9 Nov 2025, Berlin
Lu Fang: Her chip offers extreme energy efficiency while increasing computational power for AI.
Artificial intelligence is hungry for energy and computing power. Lu Fang, Professor at Tsinghua University, has taken a bold step toward solving this challenge. With her “Taichi” chip, she has developed the first architecture for photonic computing that combines extreme energy efficiency with massive processing power for AI.
Instead of relying on electricity, photonic AI uses light. The result: up to one million times less energy consumption and a thousandfold increase in performance. Inspired by new insights into how the brain actually works — fast, shallow, parallel computations rather than endless deep layers — Fang’s approach lays the groundwork for a sustainable AI future. Read more.
On 9 November in Berlin, she will be honoured as a Science Breakthrough of the Year and present her work to an international audience at the Falling Walls Science Summit — one of the leading global gatherings on the future of science and research.
Press accreditation to the Science Summit is free of charge. Please use the code SCSU25_Press on our website to receive your press ticket: https://falling-walls.com/scie...
(Please note: no press activities or interviews are available for Lu Fang.)
