# Scaling Decision Optimization to 100 Million Variables and Beyond with mPDLP in NVIDIA cuOpt

DevFeed: [Scaling Decision Optimization to 100 Million Variables and Beyond with mPDLP in NVIDIA cuOpt](<https://devfeed.tech/articles/scaling-decision-optimization-to-100-million-variables-and-beyond-with-mpdlp-in-nvidia-cuopt-66967.md>)

Original publisher: [Read original article](<https://developer.nvidia.com/blog/scaling-decision-optimization-to-100-million-variables-and-beyond-with-mpdlp-in-nvidia-cuopt/>)

Author: Tanya Lenz

Published: 2026-10-07T15:45:00Z

Content type: article

Language: en

Sources: [NVIDIA Developer](<https://devfeed.tech/sources/nvidia-developer.md>), [NVIDIA Technical Blog](<https://devfeed.tech/sources/nvidia-technical-blog.md>)

Topics: [cuOpt](<https://devfeed.tech/topics/cuopt.md>), [Distributed Systems & Parallel Computing](<https://devfeed.tech/topics/distributed-systems-parallel-computing.md>), [Combinatorial optimization](<https://devfeed.tech/topics/combinatorial-optimization.md>), [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [Operations research and optimization](<https://devfeed.tech/topics/operations-research-and-optimization.md>), [NCCL](<https://devfeed.tech/topics/nccl.md>)

Tags: [agentic-ai-generative-ai](<https://devfeed.tech/tags/agentic-ai-generative-ai.md>), [algorithms](<https://devfeed.tech/tags/algorithms.md>), [algorithms-numerical-techniques](<https://devfeed.tech/tags/algorithms-numerical-techniques.md>), [bandwidth](<https://devfeed.tech/tags/bandwidth.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [broadcast](<https://devfeed.tech/tags/broadcast.md>), [compute](<https://devfeed.tech/tags/compute.md>), [cuopt](<https://devfeed.tech/tags/cuopt.md>), [developer-tools-techniques](<https://devfeed.tech/tags/developer-tools-techniques.md>), [distributed](<https://devfeed.tech/tags/distributed.md>), [linear-programming](<https://devfeed.tech/tags/linear-programming.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [simulation-modeling-design](<https://devfeed.tech/tags/simulation-modeling-design.md>), [top-stories](<https://devfeed.tech/tags/top-stories.md>)

## AI overview

NVIDIA describes cuOpt mPDLP, a multi-GPU solver for large linear programming problems. It partitions sparse matrices to keep related work local and reduce communication across GPUs. Benchmarks show that speedups depend on problem size and sparsity: mPDLP improves on single-GPU cuOpt and often on D-PDLP for large instances, though it is slower on three ultra-large benchmark cases. Partner examples report speedups for supply chain and energy planning models.

## Source excerpt

Supply chain problems are expanding across more SKUs, lanes, and constraints than ever before, while energy grids are balancing more distributed sources in real...