{"id":"arxiv/2610.21890","url":"https://arxiv.org/abs/2610.21890","tags":["math.OC"],"score":0,"title":"GPU Parallel Pareto Local Search for Multi-objective QUBO","domain":"math.OC","source":"arxiv","authors":["Stephanie Riedmüller","Thorsten Koch"],"reviews":0,"subject":"math.OC","summary":"We revisit Pareto Local Search (PLS) for multi-objective quadratic unconstrained binary optimization (QUBO) problems. Multi-objective QUBO formulations have gained increasing attention as a universal modeling method and natural input for quantum optimization. While exact methods exist, their unconstrained binary structure makes them well-suited for heuristics such as PLS. At the same time, GPUs offer a powerful classical alternative to quantum-inspired optimization. We develop an efficient GPU-based parallel implementation of PLS for multi-objective QUBO problems. Our approach combines…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-15","reactions":{}}