This project focuses on the development and promotion of Digital Annealing (DA) and GPU-based annealing technologies, combining hardware acceleration with software innovation. On the hardware side, DA and GPU annealing are applied to combinatorial optimization, scheduling, experimental design, drug molecule discovery, photomask inverse design, and MicroLED process optimization. On the software side, the work centers on QUBO model construction and transformation, AI-enhanced solving strategies, and GPU parallel performance improvement.

The project will also establish a localized annealing platform to reduce cost, enhance flexibility, and ensure privacy, thereby supporting both academic research and industrial adoption. Through this integrat

In drug design, traditional CPU and GPU architectures are unable to efficiently handle the enormous computational challenge of estimating binding free energy across up to 10⁶⁰ possible molecular combinations. Research conducted by Dr. Jung-Hsiung Lin, Deputy Director of the Biomedical Translation Research Center at Academia Sinica, has demonstrated that Digital Annealing (DA) provides a novel and effective solution. By rapidly exploring vast solution spaces, DA can accurately predict ligand binding positions and significantly improve computational efficiency. This approach not only offers a new pathway for tackling problems previously considered intractable but also provides a practical computational framework that complements conventional 

Digital Annealing (DA) accelerates drug discovery by improving molecular screening and binding energy evaluation, thus enhancing efficiency and reducing experimental costs. It is a valuable tool for advancing new drug development. In combinatorial optimization, DA addresses key problems: solving scheduling for better workforce and resource allocation, optimizing routing for logistics and transportation, and tackling knapsack problems for portfolio and asset management. Beyond healthcare, DA is applicable in semiconductors for photomask inverse design, in display technologies for optimizing MicroLED fabrication, and in manufacturing for smarter scheduling and higher capacity utilization. These diverse applications highlight DA’s broad cross-