# NVIDIA/cuopt

GPU accelerated decision optimization

Repository: https://github.com/NVIDIA/cuopt
Canonical: https://ross.abutalabs.com/products/cuopt
Homepage: https://docs.nvidia.com/cuopt/user-guide/latest/introduction.html
Language: Cuda
License: Apache-2.0
License Family: permissive
Topics: gpu, linear-programming, optimization, cuda
Last push: 2026-09-02T22:55:17+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 84, longevity 36
- inputs: {"age_days": 512, "days_push": 0, "days_rel": 27, "gap_med": 57, "n_releases_24m": 4}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1032, forks 225 (observed 2026-09-03T02:15:06.811026+00:00)

## What it is
NVIDIA cuOpt is a GPU-accelerated optimization engine for linear programming, quadratic programming, mixed integer programming, and vehicle routing problems, built on a C++ core with C, Python, CLI, and server APIs. It scales to problems with millions of variables and constraints and integrates with modeling languages like CVXPY, Pyomo, JuMP, AMPL, GAMS, and PuLP.

## Use cases
- solve large-scale linear programming problems on GPU
- vehicle routing optimization for delivery fleets
- solve mixed integer programming problems fast
- accelerate CVXPY or Pyomo optimization workflows
- route planning for trucks with time windows and capacity constraints
- solve quadratic programming problems at scale
- deploy optimization solver as a self-hosted microservice

## When to choose
- you need near real-time solutions for very large LP/QP/MIP problems
- you have NVIDIA GPUs and want GPU-accelerated operations research
- you need vehicle routing (TSP/VRP/PDP) with heterogeneous fleets and time windows
- you want to keep existing modeling-language workflows but speed up solving

## When to avoid
- you have no NVIDIA GPU hardware
- you need proven optimality guarantees for MIP, which is still beta
- your problems are small enough that CPU solvers suffice
- you need QCQP/SOCP support outside CVXPY, which is limited

## Facets
- artifact type: library
- maturity: active
- function: math, sdk, cli, gpu-computing
- domain: logistics, gpu-computing, data-science, performance
- platform: python, cpp, cloud, self-hosted
- tags: optimization, linear-programming, vehicle-routing, operations-research, mixed-integer-programming, cuda, solver, convex-optimization, nvidia, algorithms, linux, docker, gpu

## Member repositories
- NVIDIA/cuopt (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:06.811026+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:07:37.381623+00:00, confidence not recorded.
  - readme: https://github.com/NVIDIA/cuopt (fetched 2026-09-03T02:15:06.811026+00:00, sha b0cbf3a79d17)
  - homepage: https://docs.nvidia.com/cuopt/user-guide/latest/introduction.html (fetched 2026-08-29T13:07:18.963019+00:00, sha 4d34cc94b24f)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/install.html (fetched 2026-08-29T13:07:18.972561+00:00, sha c0a63b2ff96a)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/routing-features.html (fetched 2026-08-29T13:07:18.975670+00:00, sha 705ecf22eaab)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/convex-features.html (fetched 2026-08-29T13:07:18.979666+00:00, sha 19997b9a51b9)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/milp-features.html (fetched 2026-08-29T13:07:18.982379+00:00, sha 04f610d1845e)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/cuopt-c/quick-start.html (fetched 2026-08-29T13:07:18.984518+00:00, sha b184655a00c4)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/cuopt-python/quick-start.html (fetched 2026-08-29T13:07:18.986266+00:00, sha 9c28892fed6d)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/cuopt-server/quick-start.html (fetched 2026-08-29T13:07:18.988025+00:00, sha 567343ed076a)
  - site_page: https://docs.nvidia.com/cuopt/user-guide/latest/cuopt-server/csp-guides/csp-azure.html (fetched 2026-08-29T13:07:18.990084+00:00, sha 6964be623260)
- Data as of 2026-08-30T08:39:29.467469+00:00.