[1] T. Achterberg and T. Berthold. Improving the feasibility pump. Discrete Optimization 4(1):77–86, 2007.Crossref, Google Scholar
[2] AMPL. Breaking barriers in optimization: AMPL’s Early results with NVIDIA cuOpt. Accessed July 9, 2026, https://ampl.com/blog/breaking-barriers-in-optimization-ampls-early-results-with-nvidia-cuopt/, 2025.Google Scholar
[3] D. Applegate, M. Díaz, O. Hinder, H. Lu, M. Lubin, B. O’Donoghue, and W. Schudy. Practical large-scale linear programming using primal-dual hybrid gradient. M. Ranzato, A. Beygelzimer, Y. Dauphin, P. Liang, and J. W. Vaughan, editors. Advances in Neural Information Processing Systems, volume 34. Neural Information Processing Systems Foundation Inc., San Diego, 18131–18143, 2021.Google Scholar
[4] D. Applegate, M. Diaz, O. Hinder, H. Lu, M. Lubin, B. O’Donoghue, and W. Schudy. PDLP: A practical first-order method for large-scale linear programming. Preprint, submitted January 13, https://arxiv.org/abs/2501.07018, 2025.Google Scholar
[5] Artelys. Supercharging optimisation: How Artelys, FICO and NVIDIA cuOpt join efforts to scale up energy system optimisation. Accessed July 9, 2026, https://www.artelys.com/news/artelys-nvidia-fico-collaboration-power-system-optimisation/, 2025.Google Scholar
[6] L. Bertacco, M. Fischetti, and A. Lodi. A feasibility pump heuristic for general mixed-integer problems. Discrete Optimization 4(1):63–76, 2007.Crossref, Google Scholar
[7] T. Berthold, A. Lodi, and D. Salvagnin. Ten years of feasibility pump, and counting. EURO Journal on Computational Optimization 7(1):1–14, 2019.Crossref, Google Scholar
[8] N. Blin, S. Gualandi, C. Maes, A. Lodi, and B. Stellato. Batched first-order methods for parallel LP solving in MIP. Preprint, submitted January 29, https://arxiv.org/abs/2601.21990v1, 2026.Google Scholar
[9] Cardinal Operations. Accessed July 9, 2026, https://www.shanshu.ai/copt, 2026.Google Scholar
[10] A. Chambolle and T. Pock. A first-order primal-dual algorithm for convex problems with applications to imaging. Journal of Mathematical Imaging and Vision 40(1):120–145, 2011.Crossref, Google Scholar
[11] K. Chen, D. Sun, Y. Yuan, G. Zhang, and X. Zhao. HPR-LP: An implementation of an HPR method for solving linear programming. Preprint, submitted August 22, https://arxiv.org/abs/2408.12179, 2024.Google Scholar
[12] K. Chen, D. Sun, Y. Yuan, G. Zhang, and X. Zhao. HPR-QP: A dual Halpern Peaceman-Rachford method for solving large-scale convex composite quadratic programming. Preprint, submitted July 3, https://arxiv.org/abs/2507.02470, 2025.Google Scholar
[13] Y. Chen, D. Tse, P. Nobel, P. Goulart, and S. Boyd. CuClarabel: GPU acceleration for a conic optimization solver. Preprint, submitted December 6, https://arxiv.org/abs/2412.19027, 2024.Google Scholar
[14] COIN-OR. cuOpt: GPU-accelerated optimization. GitHub repository, mirror of the NVIDIA cuOpt project. Accessed July 9, 2026, https://github.com/coin-or/cuopt/, 2026.Google Scholar
[15] A. Corduk, P. Sielski, and M. Chung. Record-breaking NVIDIA cuOpt algorithms deliver route optimization solutions 100x faster. Accessed July 9, 2026, https://developer.nvidia.com/blog/record-breaking-nvidia-cuopt-algorithms-deliver-route-optimization-solutions-100x-faster/, 2024.Google Scholar
[16] A. Corduk, P. Sielski, A. Boucher, and K. Aatish. GPU-accelerated primal heuristics for mixed integer programming. Preprint, submitted October 23, https://arxiv.org/abs/2510.20499, 2025.Google Scholar
[17] Crew AI. Accessed July 9, 2026, https://crewai.com/, 2026.Google Scholar
[18] G. B. Dantzig and J. H. Ramser. The truck dispatching problem. Management Science 6(1):80–91, 1959.Link, Google Scholar
[19] S. S. Dey, A. Iroume, M. Molinaro, and D. Salvagnin. Improving the randomization step in feasibility pump. SIAM Journal on Optimization 28(1):355–378, 2018.Crossref, Google Scholar
[20] FICO. Accessed July 9, 2026, https://www.fico.com/en/products/fico-xpress-optimization, 2026.Google Scholar
[21] M. Fischetti and D. Salvagnin. Feasibility pump 2.0. Mathematical Programming Computation 1(2):201–222, 2009.Crossref, Google Scholar
[22] M. Fischetti, F. Glover, and A. Lodi. The feasibility pump. Mathematical Programming 104:91–104, 2005.Crossref, Google Scholar
[23] GAMS. Expangind the focus: Introducing the MIPfeas benchmark. Accessed July 9, 2026, https://www.gams.com/blog/2026/03/expanding-the-focus-introducing-the-mipfeas-benchmark/, 2026.Google Scholar
[24] D. Ge, C. Wang, Z. Xiong, and Y. Ye. From an interior point to a corner point: Smart crossover. INFORMS Journal on Computing 37(6):1670–1688, 2025.Google Scholar
[25] Gurobi Optimization. The engine behind your most complex decisions. Accessed July 9, 2026, https://www.gurobi.com/, 2026.Google Scholar
[26] P. Huo, F. Huang, and I. Boier. Accelerating real-time financial decisions with quantitative portfolio optimization. Accessed July 9, 2026, https://developer.nvidia.com/blog/accelerating-real-time-financial-decisions-with-quantitative-portfolio-optimization/, 2025.Google Scholar
[27] INFORMS. 2026 Edelman Award Presentation: NVIDIA. Accessed July 9, 2026, https://www.youtube.com/watch?v=23_JBl0xaS4, 2026.Google Scholar
[28] Kinaxis. Kinaxis advances large-scale supply chain optimization with NVIDIA AI. Accessed July 9, 2026, https://investors.kinaxis.com/news-releases/news-release-details/2026/Kinaxis-Advances-Large-Scale-Supply-Chain-Optimization-with-NVIDIA-AI/default.aspx, 2026.Google Scholar
[29] LangChain. Powering the agent development lifecycle. Accessed July 9, 2026, https://www.langchain.com/, 2026.Google Scholar
[30] LangGraph. Balance agent control with agency. Accessed July 9, 2026, https://www.langchain.com/langgraph, 2026.Google Scholar
[31] Z. Lei, J.-K. Hao, and Q. Wu. Speeding up local optimization in vehicle routing with tensor-based GPU acceleration. Preprint, submitted June 20, https://arxiv.org/abs/2506.17357, 2025.Google Scholar
[32] P. Lin, M. Zou, and S. Cai. An efficient local search solver for mixed integer programming. P. Shaw, editor. Proceedings of the 30th International Conference on Principles and Practice of Constraint Programming, volume 307. Schloss Dagstuhl – Leibniz-Zentrum für Informatik, Wadern, Germany, 1–19, 2024.Google Scholar
[33] Y. Liu. cuGenOpt: A GPU-accelerated general-purpose metaheuristic framework for combinatorial optimization. Preprint, submitted March 19, https://arxiv.org/abs/2603.19163, 2026.Google Scholar
[34] T. Liu and H. Lu. A new crossover algorithm for LP inspired by the spiral dynamic of PDHG. Preprint, submitted September 23, https://arxiv.org/abs/2409.14715, 2024.Google Scholar
[35] H. Lu. cuPDLPx. Accessed July 9, 2026, https://github.com/MIT-Lu-Lab/cuPDLPx, 2026.Google Scholar
[36] H. Lu and J. Yang. Restarted Halpern PDHG for linear programming. Preprint, submitted July 23, https://arxiv.org/abs/2407.16144, 2024.Google Scholar
[37] H. Lu and J. Yang. cuPDLP.jl: A GPU implementation of restarted primal-dual hybrid gradient for linear programming in Julia. Operations Research 73(6):3440–3452, 2025.Link, Google Scholar
[38] H. Lu and J. Yang. An overview of GPU-based first-order methods for linear programming and extensions. Preprint, submitted June 2, https://arxiv.org/abs/2506.02174, 2025.Google Scholar
[39] H. Lu, Z. Peng, and J. Yang. MPAX: Mathematical programming in JAX. Preprint, submitted December 12, https://arxiv.org/abs/2412.09734, 2024.Google Scholar
[40] H. Lu, Z. Peng, and J. Yang. cuPDLPx: A further enhanced gpu-based first-order solver for linear programming. Preprint, submitted July 18, https://arxiv.org/abs/2507.14051, 2025.Google Scholar
[41] B. Luteberget and G. Sartor. Feasibility jump: An LP-free Lagrangian MIP heuristic. Mathematical Programming Computation 15(2):365–388, 2023.Crossref, Google Scholar
[42] C. Maes. Solve linear programs using the GPU-accelerated barrier method in NVIDIA cuOpt. Accessed July 9, 2026, https://developer.nvidia.com/blog/solve-linear-programs-using-the-gpu-accelerated-barrier-method-in-nvidia-cuopt/, 2025.Google Scholar
[43] S. Mehrotra. On the implementation of a primal-dual interior point method. SIAM Journal on Optimization 2(4):575–601, 1992.Crossref, Google Scholar
[44] Microsoft. AutoGen. Accessed July 9, 2026, https://microsoft.github.io/autogen/stable//index.html, 2026.Google Scholar
[45] Microsoft. Optiguide. Accessed July 9, 2026, https://github.com/microsoft/optiguide, 2026.Google Scholar
[46] H. D. Mittelmann. Mittelmann benchmarks. Accessed June 9, 2026, https://plato.asu.edu/bench.html, 2026.Google Scholar
[47] H. D. Mittelmann. Mittelmann benchmarks, LPfeas category. Accessed June 9, 2026, https://plato.asu.edu/ftp/lpfeas.html, 2026.Google Scholar
[48] H. D. Mittelmann. Mittelmann benchmarks, MIPfeas category. Accessed June 9, 2026, https://plato.asu.edu/ftp/mipfeas.html, 2026.Google Scholar
[49] NVIDIA. CHOLMOD. NVIDIA developer. Accessed July 9, 2026, https://developer.nvidia.com/cholmod, 2014.Google Scholar
[50] NVIDIA. cuOpt. Accessed July 9, 2026, https://github.com/NVIDIA/cuopt, 2026.Google Scholar
[51] NVIDIA. cuOpt agent. Accessed July 9, 2026, https://github.com/NVIDIA/cuopt-examples/tree/main/cuopt-agent, 2026.Google Scholar
[52] NVIDIA. cuOpt examples. Accessed July 9, 2026, https://github.com/NVIDIA/cuopt-examples, 2026.Google Scholar
[53] NVIDIA. cuOpt examples. Accessed July 9, 2026, https://github.com/NVIDIA/cuopt-examples/blob/main/diet_optimization/diet_optimization_lp.ipynb, 2026.Google Scholar
[54] NVIDIA. cuOpt examples. Accessed July 9, 2026, https://github.com/NVIDIA/cuopt-examples/blob/main/emergency_management/capacitated_emergency_shelter.ipynb, 2026.Google Scholar
[55] NVIDIA. cuOpt examples. Accessed July 9, 2026, https://github.com/NVIDIA/cuopt-examples/blob/main/portfolio_optimization/QP_portfolio_optimization.ipynb, 2026.Google Scholar
[56] NVIDIA. cuOpt examples. Accessed July 9, 2026, https://github.com/NVIDIA/cuopt-examples/blob/main/last_mile_delivery/cvrp_daily_deliveries.ipynb, 2026.Google Scholar
[57] NVIDIA. NeMo relay. Accessed July 9, 2026, https://docs.nvidia.com/nemo/relay/about-nemo-relay/overview, 2026.Google Scholar
[58] NVIDIA. NVLabs: EARLI. Accessed July 9, 2026, https://github.com/NVlabs/EARLI, 2026.Google Scholar
[59] M. A. O’Neil and M. Burtscher. Rethinking the parallelization of random-restart hill climbing: A case study in optimizing a 2-Opt TSP solver for GPU execution. D. Kaeli and J. Cavazos, editors. Proceedings of the 8th Workshop on General Purpose Processing Using GPUs. ACM, New York, 99–108, 2015.Google Scholar
[60] OSQP. Python interface for CUDA implementation of OSQP. Accessed July 9, 2026, https://github.com/osqp/cuosqp, 2026.Google Scholar
[61] E. Rothberg. New options for solving giant LPs. Accessed July 9, 2026, https://youtu.be/B9rKByy2ym4 and https://cdn.gurobi.com/wp-content/uploads/New-Options-for-Solving-Giant-LPs.pdf, 2024.Google Scholar
[62] E. Rothberg. Backing PDHG into a corner. Preprint, submitted November 17, https://arxiv.org/abs/2511.13894, 2025.Google Scholar
[63] E. Rothberg. Concurrent crossover for PDHG. Preprint, submitted October 28, https://arxiv.org/abs/2510.24429, 2025.Google Scholar
[64] E. Rothberg. Empirical asymptotic runtime analysis of linear programming algorithms. Preprint, submitted April 17, https://arxiv.org/abs/2604.16192, 2026.Google Scholar
[65] D. Salvagnin, R. Roberti, and M. Fischetti. A fix-propagate-repair heuristic for mixed integer programming. Mathematical Programming Computation 17(1):111–139, 2025.Crossref, Google Scholar
[66] M. Savelsbergh. Preprocessing and probing techniques for mixed integer programming problems. ORSA Journal on Computing 6(4):317–454, 1994.Link, Google Scholar
[67] M. Schubiger, G. Banjac, and J. Lygeros. GPU acceleration of ADMM for large-scale quadratic programming. Journal of Parallel and Distributed Computing 144:55–67, 2020.Crossref, Google Scholar
[68] P. Sielski, A. Corduk, and A. Boucher. Learn how NVIDIA cuOpt accelerates mixed integer optimization using primal heuristics. Accessed July 9, 2026, https://developer.nvidia.com/blog/learn-how-nvidia-cuopt-accelerates-mixed-integer-optimization-using-primal-heuristics/, 2026.Google Scholar
[69] P. Sielski, A. Çördük, H. Linsenmaier, and A. Fender. A GPU parallel algorithm for finding a negative subset disjoint cycle in a graph. Proceedings of the 2023 IEEE High Performance Extreme Computing Conference. IEEE, Boston, 1–7, 2023.Google Scholar
[70] SINTEF. Li & Lim benchmark. Accessed July 9, 2026, https://www.sintef.no/projectweb/top/pdptw/li-lim-benchmark/, 2026.Google Scholar
[71] SINTEF. Gehring & Homberger benchmark. Accessed July 9 2026, https://www.sintef.no/projectweb/top/vrptw/homberger-benchmark/, 2026.Google Scholar
[72] R. Thind, Y. Sun, L. Liang, and H. Yang. OptimAI: Optimization from natural language using LLM-powered AI agents. Preprint, submitted April 23, https://arxiv.org/abs/2504.16918, 2025.Google Scholar
[73] T. Van Luong, N. Melab, and E.-G. Talbi. GPU-based multi-start local search algorithms. C. A. C. Coello, editor. Learning and Intelligent Optimization, Lecture Notes in Computer Science, volume 6683. Springer, Berlin, 321–335, 2011.Google Scholar
[74] J. Zhang, W. Wang, S. Guo, L. Wang, F. Lin, C. Yang, and W. Yin. Solving general natural-language-description optimization problems with large language models. Y. Yang, A. Davani, A. Sil, and A. Kumar, editors. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, volume 6. Association for Computational Linguistics, Mexico City, 483–490, 2024.Google Scholar
[75] J. Zhao, L. Yang, H. Zhang, Q. He, and T. Ding. From sequential to parallel: Reformulating dynamic programming as GPU kernels for large-scale stochastic combinatorial optimization. Proceedings of the International Conference on Learning Representations. ICLR, 2026.Google Scholar

cover image INFORMS TutORials in Operations Research

Operations Research and Artificial Intelligence: Better Together

November 2026

Article Information

Metrics

Information

Published Online:November 01, 2026

Cite as

Nicolas Blin, Burcin Bozkaya, Akif Coerduek, Christopher Maes (2026) GPU-Accelerated Decision Optimization. INFORMS TutORials in Operations Research null(null):377-400.

https://doi.org/10.1287/educ.2026.0306

Keywords

PDF download