UNSW AI System Speeds Up Semiconductor Discovery

Researchers at the University of New South Wales (UNSW) launched an AI-driven system on April 21, 2026, to accelerate semiconductor discovery ^[1]. The workflow identifies two-dimensional hybrid perovskites for high-efficiency solar cells and LEDs ^[2]. This method helps scientists find promising materials in a fraction of the usual time ^[3].

The AI filters millions of molecular combinations by working backward from specific electrical requirements ^[1]. By focusing on desired performance, the system narrows down candidates for laboratory testing ^[2]. This innovation simplifies the search for next-generation electronic materials ^[3].