Derives linear sample complexity for PDHG parameters and polynomial sample complexity for full PDLP hyperparameters using data-driven algorithm design.
cupdlpx: A further enhanced gpu-based first-order solver for linear programming
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
years
2026 6roles
method 1polarities
use method 1representative citing papers
CHAP's cross-platform portfolio finds feasible solutions for 47 of 50 MIP benchmark instances in five minutes, beating Gurobi (44) and cuOpt (43) by coordinating GPU tabu search with CPU fix-and-propagate and feasibility pump via a shared pool.
A set of simple low-cost presolve rules captures most of Gurobi's reduction and yields end-to-end speedups for GPU first-order LP solvers.
D-PDLP is the first distributed multi-GPU framework for PDLP that uses 2D grid partitioning of the constraint matrix plus nonzero-aware and random-permutation strategies to scale PDHG iterations with low overhead and full FP64 accuracy.
A PyTorch-based multi-GPU LP solver using column-sharded parallelism, fused kernels, and ridge regularization claims order-of-magnitude speedups and near-linear scaling on GPU clusters for large matching problems.
Regression models fit observed LP solver runtimes well within instance classes, but asymptotic growth rates differ substantially across simplex, interior-point, and PDHG methods.
citing papers explorer
-
Parameter Tuning with Generalization Guarantees for GPU-Accelerated Linear Programming
Derives linear sample complexity for PDHG parameters and polynomial sample complexity for full PDLP hyperparameters using data-driven algorithm design.
-
CHAP: A Hybrid GPU-CPU Heuristic for MIP
CHAP's cross-platform portfolio finds feasible solutions for 47 of 50 MIP benchmark instances in five minutes, beating Gurobi (44) and cuOpt (43) by coordinating GPU tabu search with CPU fix-and-propagate and feasibility pump via a shared pool.
-
Presolving for GPU-Accelerated First-Order LP Solvers
A set of simple low-cost presolve rules captures most of Gurobi's reduction and yields end-to-end speedups for GPU first-order LP solvers.
-
D-PDLP: Scaling PDLP to Distributed Multi-GPU Systems
D-PDLP is the first distributed multi-GPU framework for PDLP that uses 2D grid partitioning of the constraint matrix plus nonzero-aware and random-permutation strategies to scale PDHG iterations with low overhead and full FP64 accuracy.
-
Large-Scale Regularized Matching on GPU Clusters
A PyTorch-based multi-GPU LP solver using column-sharded parallelism, fused kernels, and ridge regularization claims order-of-magnitude speedups and near-linear scaling on GPU clusters for large matching problems.
-
Empirical Asymptotic Runtime Analysis of Linear Programming Algorithms
Regression models fit observed LP solver runtimes well within instance classes, but asymptotic growth rates differ substantially across simplex, interior-point, and PDHG methods.