{"id":"arxiv/2610.38706","url":"https://arxiv.org/abs/2610.38706","tags":["math.OC","cs.MS"],"score":0,"title":"cuPDQPx: A GPU-Accelerated Restarted Halpern PDHG for Large-Scale Convex Quadratic Programming","domain":"math.OC","source":"arxiv","authors":["Haihao Lu","Zedong Peng","Jinwen Yang"],"reviews":0,"subject":"math.OC","summary":"We present cuPDQPx, a GPU-accelerated, factorization-free first-order solver for large-scale convex quadratic programming (QP). Building on the restarted reflected Halpern primal-dual framework of cuPDLPx, we introduce an auxiliary-variable splitting that yields explicit updates without linearizing the quadratic objective or solving inner quadratic subproblems. A preconditioned proximal point interpretation yields both sublinear and linear fixed-point residual convergence guarantees. For practical efficiency, cuPDQPx integrates diagonal preconditioning, step-size selection, adaptive restart,…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-28","reactions":{}}