The Research Centre for Intelligent Operations Research (IOR) at the Department of Applied Mathematics (AMA), PolyU, is proud to announce the open-source release of two powerful GPU solvers, HPR-QP and HPR-LP, developed in collaboration with Beijing University of Technology. The PolyU team was led by Professor Sun Defeng, Head of Department, Chair Professor of Applied Optimisation and Operations Research, and Director of IOR.
HPR-QP and HPR-LP are Julia-based implementations of the Halpern Peaceman–Rachford (HPR) method, designed to efficiently solve large-scale convex composite quadratic programming (CCQP) and linear programming (LP) problems on GPUs. Now freely available to the global research community on the open-source platform GitHub, these solvers are poised to benefit both academia and industry by supporting advanced computational optimisation and a wide range of real-world applications.
About the Research Centre for Intelligent Operations Research (IOR)
Established in August 2024, IOR is dedicated to pioneering research in operations research, optimisation algorithm design, solver development, and their applications, leveraging state-of-the-art AI methodologies. Intelligent operations research is a dynamic and sophisticated scientific domain focused on developing cutting-edge algorithms and models to automate and enhance decision-making processes, and to uncover optimal solutions within highly complex systems.
The Centre is committed to advancing knowledge and innovation in intelligent optimisation and operations research, while fostering a vibrant and responsible research community that delivers practical benefits to society at large.
For more details and access to the solvers, please visit:
HPR-QP: A GPU Solver for Convex Composite Quadratic Programming in Julia
HPR-LP: A GPU Solver for Linear Programming
