GAMS solver link for NVIDIA cuOpt solver

This project builds and packages the GAMS and GAMSPy solver link for the NVIDIA cuOpt solver.

You can get more details and tips by reading the blog post "GPU-Accelerated Optimization with GAMS and NVIDIA cuOpt".

Requirements

• Operating System: Linux, Windows 11 through WSL2
• GAMS: Version 49 or newer
• GAMSPy: Version 1.12.1 or newer
• NVIDIA GPU: Volta architecture or better
• CUDA Runtime Libraries: 12 or 13

Getting started / installation

• Make sure CUDA runtime is installed
• Download and unpack cuopt-link-release-cu12.zip or cuopt-link-release-cu13.zip (for CUDA 12 and 13 respectively) from the releases page:
  • Unpack the contents of cuopt-link-release-cu*.zip into your GAMS system directory. For GAMSPy, you can find out your system directory by running gamspy show base. So for example you can run unzip -o cuopt-link-release-cu*.zip -d $(gamspy show base).
  • Caution: This will overwrite any existing gamsconfig.yaml file in that directory. The contained gamsconfig.yaml contains a solverConfig section to make cuOpt available to GAMS.

The neccessary files from the CUDA 12 or 13 runtime can also be downloaded as convenient archive cu12-runtime.zip or cu13-runtime.zip from the releases page.

Test the setup

Get an example model and explicitly choose cuopt as lp or mip solver:

gamslib trnsport
gams trnsport lp cuopt

Examples

• examples/trnsport_cuopt.ipynb for CUDA 12
• examples/trnsport_cuopt.ipynb for CUDA 13