# NVIDIA cuOpt **URL:** https://forums.developer.nvidia.com/c/ai-data-science/nvidia-cuopt/514.md [Latest](https://forums.developer.nvidia.com/latest.md) · [Categories](https://forums.developer.nvidia.com/categories.md) · [Tags](https://forums.developer.nvidia.com/tags.md) --- ## [Could not select device driver "" with capabilities \[\[gpu\]\]](https://forums.developer.nvidia.com/t/could-not-select-device-driver-with-capabilities-gpu/214344)
**Author:** [@ryanmelnick](https://forums.developer.nvidia.com/u/ryanmelnick)\ **Replies:** 3\ **Last updated:** [January 26, 2026, 2:40am UTC](https://forums.developer.nvidia.com/t/could-not-select-device-driver-with-capabilities-gpu/214344 "2026-01-26T02:40:48Z")
I am running rootless docker with Ubuntu 20.04 cuda 10.1.243 Docker 20.10.14 nvidia-docker2 I run the follwing with the error… docker run --entrypoint /bin/bash -it --gpus all --rm nvcr.io/ea-reopt-member-zone/ea-cu… --- ## [Update: cuOpt Community Support Moving to GitHub](https://forums.developer.nvidia.com/t/update-cuopt-community-support-moving-to-github/347673)
**Author:** [@rnarasimha1](https://forums.developer.nvidia.com/u/rnarasimha1)\ **Replies:** 0\ **Last updated:** [October 14, 2025, 10:47pm UTC](https://forums.developer.nvidia.com/t/update-cuopt-community-support-moving-to-github/347673 "2025-10-14T22:47:02Z")
📢 Update: cuOpt Community Support Moving to GitHub Since NVIDIA cuOpt™ is now open source, we’ll continue supporting all cuOpt-related questions and discussions in our newly created cuOpt GitHub Community. We invite yo… --- ## [The cuOpt API is crashing with a 500 Internal Server Error and a CUDA illegal memory access deta](https://forums.developer.nvidia.com/t/the-cuopt-api-is-crashing-with-a-500-internal-server-error-and-a-cuda-illegal-memory-access-deta/344593)
**Author:** [@SiddheshramK](https://forums.developer.nvidia.com/u/SiddheshramK)\ **Replies:** 5\ **Last updated:** [September 17, 2025, 8:16am UTC](https://forums.developer.nvidia.com/t/the-cuopt-api-is-crashing-with-a-500-internal-server-error-and-a-cuda-illegal-memory-access-deta/344593 "2025-09-17T08:16:33Z")
The cuOpt API is crashing with a 500 Internal Server Error and a CUDA illegal memory access detail. I am using the cuOpt API to solve a vehicle routing problem. When I send a request using my dataset, the server returns… --- ## [Numerical stability issues](https://forums.developer.nvidia.com/t/numerical-stability-issues/344048)
**Author:** [@chintasunny](https://forums.developer.nvidia.com/u/chintasunny)\ **Replies:** 4\ **Last updated:** [September 5, 2025, 3:40pm UTC](https://forums.developer.nvidia.com/t/numerical-stability-issues/344048 "2025-09-05T15:40:35Z")
I am running a small scheduling problem using cuopt. It has got large penalty values and some Big M kind of variables as well. It’s running into the below error. While the same instance is getting solved easily on coin-o… --- ## [Cuopt & pulp integration](https://forums.developer.nvidia.com/t/cuopt-pulp-integration/344020)
**Author:** [@chintasunny](https://forums.developer.nvidia.com/u/chintasunny)\ **Replies:** 3\ **Last updated:** [September 4, 2025, 3:52pm UTC](https://forums.developer.nvidia.com/t/cuopt-pulp-integration/344020 "2025-09-04T15:52:05Z")
I am able to successfully run cuopt on a gpu. When I try to run pulp integration examples, I am running into the following issue. Can you plz help me on this error /opt/conda/envs/cuopt/lib/python3.10/site-packages/cuop… --- ## [\[BUG\]: std::bad\_alloc: out\_of\_memory: CUDA error](https://forums.developer.nvidia.com/t/bug-std-out-of-memory-cuda-error/342885)
**Author:** [@alberto.manzini](https://forums.developer.nvidia.com/u/alberto.manzini)\ **Replies:** 4\ **Last updated:** [August 25, 2025, 1:06pm UTC](https://forums.developer.nvidia.com/t/bug-std-out-of-memory-cuda-error/342885 "2025-08-25T13:06:20Z")
Running this example LP and MILP Examples — NVIDIA cuOpt (25.08) I get this error Setting parameter time\_limit to 6.000000e+01 Problem status: NoTermination Solving a problem with 2 constraints 2 variables (2 integer… --- ## [Cuopt\_cli --method 0,1,2 using saming level of CPU/GPU resources](https://forums.developer.nvidia.com/t/cuopt-cli-method-0-1-2-using-saming-level-of-cpu-gpu-resources/341145)
**Author:** [@westlywschmidt](https://forums.developer.nvidia.com/u/westlywschmidt)\ **Replies:** 1\ **Last updated:** [August 11, 2025, 5:53pm UTC](https://forums.developer.nvidia.com/t/cuopt-cli-method-0-1-2-using-saming-level-of-cpu-gpu-resources/341145 "2025-08-11T17:53:43Z")
I am trying to solve a LP problem with 4391 constraints 4428 variables (2223 integers) and 14233 nonzeros. The GPU is an ASUS NVIDIA 5090 OC with default settings and an AMD 9950X3D CPU running on an Arch based distro -… --- ## [Is it possible to disable strong branching in cuOpt?](https://forums.developer.nvidia.com/t/is-it-possible-to-disable-strong-branching-in-cuopt/337863)
**Author:** [@nconnolly](https://forums.developer.nvidia.com/u/nconnolly)\ **Replies:** 5\ **Last updated:** [July 7, 2025, 8:36pm UTC](https://forums.developer.nvidia.com/t/is-it-possible-to-disable-strong-branching-in-cuopt/337863 "2025-07-07T20:36:50Z")
Hi All, I am using cuOpt for an MILP problem. The problem is solved by PuLP within a few seconds with an optimal solution, while cuOpt uses up all the available time up until time\_limit, and outputs the solution with t… --- ## [Cuopt Parameter Name Incorrect](https://forums.developer.nvidia.com/t/cuopt-parameter-name-incorrect/337811)
**Author:** [@yyyxxxzzz](https://forums.developer.nvidia.com/u/yyyxxxzzz)\ **Replies:** 1\ **Last updated:** [July 2, 2025, 5:39pm UTC](https://forums.developer.nvidia.com/t/cuopt-parameter-name-incorrect/337811 "2025-07-02T17:39:13Z")
I want to set the parameterCUOPT\_MIP\_RELATIVE\_GAP. The documentation says " When setting parameters in thin client solver settings, remove CUOPT\_ from the parameter name and convert to lowercase". I used “mip\_relative\_ga… --- ## [Has https://pypi.nvidia.com/cuopt-sh-client/ been removed?](https://forums.developer.nvidia.com/t/has-https-pypi-nvidia-com-cuopt-sh-client-been-removed/337371)
**Author:** [@yyyxxxzzz](https://forums.developer.nvidia.com/u/yyyxxxzzz)\ **Replies:** 5\ **Last updated:** [July 1, 2025, 8:21pm UTC](https://forums.developer.nvidia.com/t/has-https-pypi-nvidia-com-cuopt-sh-client-been-removed/337371 "2025-07-01T20:21:33Z")
I tried to pip install cuopt-sh-client but it failed. Error message: The installation of cuopt-sh-client for version 25.5.1 failed. This is a special placeholder package which downloads a real wheel package from ht… --- ## [Unable to run cuOpt managed service and microservice](https://forums.developer.nvidia.com/t/unable-to-run-cuopt-managed-service-and-microservice/335962)
**Author:** [@abhilashcashok](https://forums.developer.nvidia.com/u/abhilashcashok)\ **Replies:** 2\ **Last updated:** [June 13, 2025, 5:54pm UTC](https://forums.developer.nvidia.com/t/unable-to-run-cuopt-managed-service-and-microservice/335962 "2025-06-13T17:54:04Z")
Hello, I have succesfully setup cuOpt via WSL in Windows 11 as per documentation - Quickstart Guide — NVIDIA cuOpt (25.05). However, I am facing multiple issues while running the cuOpt examples (eg: Simple Cost Matrix)… --- ## [How to get access to hosted cuOpt with API in inception program](https://forums.developer.nvidia.com/t/how-to-get-access-to-hosted-cuopt-with-api-in-inception-program/333807)
**Author:** [@asaf.horvitz](https://forums.developer.nvidia.com/u/asaf.horvitz)\ **Replies:** 8\ **Last updated:** [May 29, 2025, 4:21pm UTC](https://forums.developer.nvidia.com/t/how-to-get-access-to-hosted-cuopt-with-api-in-inception-program/333807 "2025-05-29T16:21:07Z")
I am in the NVIDIA Inception Program. I want to get access to hosted cuOpt with API in inception program. How it is done ? --- ## [Is there a way to use CuOPT in C# with Visual Studio?](https://forums.developer.nvidia.com/t/is-there-a-way-to-use-cuopt-in-c-with-visual-studio/330029)
**Author:** [@skwix7905](https://forums.developer.nvidia.com/u/skwix7905)\ **Replies:** 2\ **Last updated:** [April 22, 2025, 2:25pm UTC](https://forums.developer.nvidia.com/t/is-there-a-way-to-use-cuopt-in-c-with-visual-studio/330029 "2025-04-22T14:25:21Z")
I’m working on a simulation implemented in C#, and I’d like to know if there’s a way to use CuOpt with it. And if that’s not possible, I tried to use CuOpt with Python in Visual Studio Code, but I haven’t been able to. … --- ## [System Requirements - Expected for cuOpt](https://forums.developer.nvidia.com/t/system-requirements-expected-for-cuopt/328595)
**Author:** [@ascandaliato1](https://forums.developer.nvidia.com/u/ascandaliato1)\ **Replies:** 2\ **Last updated:** [April 9, 2025, 9:42pm UTC](https://forums.developer.nvidia.com/t/system-requirements-expected-for-cuopt/328595 "2025-04-09T21:42:59Z")
The cuOpt documentation states that A100 and H100 GPUs are required for a Self-Hosted container. I would like to know if this is likely to change. It is extremely valuable to support GeForce RTX 40 and 50 series GPUs fo… --- ## [ASAP access to cuOpt for LP and MILP](https://forums.developer.nvidia.com/t/asap-access-to-cuopt-for-lp-and-milp/328597)
**Author:** [@ascandaliato1](https://forums.developer.nvidia.com/u/ascandaliato1)\ **Replies:** 1\ **Last updated:** [April 2, 2025, 9:28am UTC](https://forums.developer.nvidia.com/t/asap-access-to-cuopt-for-lp-and-milp/328597 "2025-04-02T09:28:02Z")
If I had access to cuOpt for LP and MILP I would be using it yesterday. What is the earliest expected date that us mere mortals can start cooking with it? And, what will be the minimal / recommended system requirements? … --- ## [cuOPT routing to solve VRP simultaneously Pickup and Delivery](https://forums.developer.nvidia.com/t/cuopt-routing-to-solve-vrp-simultaneously-pickup-and-delivery/327710)
**Author:** [@yangmin1](https://forums.developer.nvidia.com/u/yangmin1)\ **Replies:** 0\ **Last updated:** [March 20, 2025, 4:23am UTC](https://forums.developer.nvidia.com/t/cuopt-routing-to-solve-vrp-simultaneously-pickup-and-delivery/327710 "2025-03-20T04:23:39Z")
May I ask whether cuOPT routing can handle VRP simultaneously Pickup and Delivery? The scene we encountered is that vehicles arrive at the store and need to unload before picking up the goods --- ## [Inquiry About Task Scheduling Technology in NVIDIA AMR Platform](https://forums.developer.nvidia.com/t/inquiry-about-task-scheduling-technology-in-nvidia-amr-platform/325377)
**Author:** [@sjb378](https://forums.developer.nvidia.com/u/sjb378)\ **Replies:** 0\ **Last updated:** [February 27, 2025, 5:56am UTC](https://forums.developer.nvidia.com/t/inquiry-about-task-scheduling-technology-in-nvidia-amr-platform/325377 "2025-02-27T05:56:51Z")
I have explored the cuOpt and Isaac Sim technologies related to NVIDIA’s AMR platform. However, I am specifically interested in task scheduling technology that allows for the selection of the most suitable robot based on… --- ## [Does cuOpt handle workload balancing among multiple robots?](https://forums.developer.nvidia.com/t/does-cuopt-handle-workload-balancing-among-multiple-robots/324973)
**Author:** [@sjb378](https://forums.developer.nvidia.com/u/sjb378)\ **Replies:** 2\ **Last updated:** [February 27, 2025, 5:52am UTC](https://forums.developer.nvidia.com/t/does-cuopt-handle-workload-balancing-among-multiple-robots/324973 "2025-02-27T05:52:26Z")
Hi, I recently came across cuOpt, and I understand that it’s designed for route optimization, even in dynamic environments like warehouses where obstacles and layouts change. That makes sense to me. However, I’m curious… --- ## [Feasible solutions could not be found. Closest infeasible solution is found with infeasibilities in: Capacity dimension](https://forums.developer.nvidia.com/t/feasible-solutions-could-not-be-found-closest-infeasible-solution-is-found-with-infeasibilities-in-capacity-dimension/325231)
**Author:** [@sebastian.boinski](https://forums.developer.nvidia.com/u/sebastian.boinski)\ **Replies:** 1\ **Last updated:** [February 26, 2025, 3:42pm UTC](https://forums.developer.nvidia.com/t/feasible-solutions-could-not-be-found-closest-infeasible-solution-is-found-with-infeasibilities-in-capacity-dimension/325231 "2025-02-26T15:42:41Z")
When trying to send data as below, I get information that the service could not process the request because the capisties conditions were not met. Does the request contain an error? I wanted to verify the Pickup and Del… --- ## [cuOpt Not Working on WSL2 + Ubuntu 22.04.5 + Docker (404 Error)](https://forums.developer.nvidia.com/t/cuopt-not-working-on-wsl2-ubuntu-22-04-5-docker-404-error/324988)
**Author:** [@sjb378](https://forums.developer.nvidia.com/u/sjb378)\ **Replies:** 1\ **Last updated:** [February 25, 2025, 3:50pm UTC](https://forums.developer.nvidia.com/t/cuopt-not-working-on-wsl2-ubuntu-22-04-5-docker-404-error/324988 "2025-02-25T15:50:26Z")
Hello, I am trying to run cuOpt on my Windows 11 machine using WSL2 with Ubuntu 22.04.5. My setup includes CUDA, and Docker runs without any issues. I have a free subscription to NVIDIA Enterprise and followed the instr… --- ## [CuOpt information and license](https://forums.developer.nvidia.com/t/cuopt-information-and-license/323783)
**Author:** [@liviolima80](https://forums.developer.nvidia.com/u/liviolima80)\ **Replies:** 5\ **Last updated:** [February 19, 2025, 4:05pm UTC](https://forums.developer.nvidia.com/t/cuopt-information-and-license/323783 "2025-02-19T16:05:01Z")
Good morning, I’m approaching to use NVIDIA libraries to solve linear optimization problems. I see that potentially this could be done with 2 approaches: CuOpt and CuSolver. CuOpt seems a more complete framework, but… --- ## [Vehicle break time windows not respected in solve](https://forums.developer.nvidia.com/t/vehicle-break-time-windows-not-respected-in-solve/316945)
**Author:** [@joakim.apelqvist](https://forums.developer.nvidia.com/u/joakim.apelqvist)\ **Replies:** 1\ **Last updated:** [January 16, 2025, 10:19am UTC](https://forums.developer.nvidia.com/t/vehicle-break-time-windows-not-respected-in-solve/316945 "2025-01-16T10:19:53Z")
I’m having an issue where most of the time, cuOpt just puts all the vehicle breaks at the same time, completely ignoring the specified time windows. Initially I had some luck where I adjusted the solver config objectives… --- ## [Docker pull returns access denied :ea-cuopt-server:\ for Server](https://forums.developer.nvidia.com/t/docker-pull-returns-access-denied-tag-for-server/211582)
**Author:** [@shivam.mehta](https://forums.developer.nvidia.com/u/shivam.mehta)\ **Replies:** 9\ **Last updated:** [November 22, 2024, 12:37am UTC](https://forums.developer.nvidia.com/t/docker-pull-returns-access-denied-tag-for-server/211582 "2024-11-22T00:37:38Z")
docker pull nvcr.io/ea-reopt-member-zone/ea-cuopt-server Using default tag: latest Error response from daemon: pull access denied for nvcr.io/ea-reopt-member-zone/ea-cuopt-server, repository does not exist or may requi… --- ## [Verbose\_mode not work? can i get more information during the solver execution](https://forums.developer.nvidia.com/t/verbose-mode-not-work-can-i-get-more-information-during-the-solver-execution/312795)
**Author:** [@yyyyy](https://forums.developer.nvidia.com/u/yyyyy)\ **Replies:** 0\ **Last updated:** [November 11, 2024, 3:33am UTC](https://forums.developer.nvidia.com/t/verbose-mode-not-work-can-i-get-more-information-during-the-solver-execution/312795 "2024-11-11T03:33:34Z")
I want to know does it’s way to display more detail execution ? I tried to set log level to debug and solver\_config.verbose\_mode = true but it seems not work. self-hosted instance with 24.07 and set log level … --- ## [I'm wondering if the cost matrix in cuOpt is evaluated based on the Euclidean distance between each point](https://forums.developer.nvidia.com/t/im-wondering-if-the-cost-matrix-in-cuopt-is-evaluated-based-on-the-euclidean-distance-between-each-point/308386)
**Author:** [@sm.seo](https://forums.developer.nvidia.com/u/sm.seo)\ **Replies:** 1\ **Last updated:** [October 1, 2024, 2:53pm UTC](https://forums.developer.nvidia.com/t/im-wondering-if-the-cost-matrix-in-cuopt-is-evaluated-based-on-the-euclidean-distance-between-each-point/308386 "2024-10-01T14:53:39Z")
Question: One of the inputs into cuOpt is the cost matrix, which is a n x n matrix showing the pairwise cost to travel from one point to another. Can we clarify how the current cost matrix is evaluated? Would the cost si… --- ## [cuOpt REST API returns "pending-evaluation"](https://forums.developer.nvidia.com/t/cuopt-rest-api-returns-pending-evaluation/306288)
**Author:** [@jozef.cepel1](https://forums.developer.nvidia.com/u/jozef.cepel1)\ **Replies:** 3\ **Last updated:** [September 16, 2024, 2:15pm UTC](https://forums.developer.nvidia.com/t/cuopt-rest-api-returns-pending-evaluation/306288 "2024-09-16T14:15:48Z")
I am working with the cuOpt REST API integrated with Azure Maps (Multi-Itinerary Optimization - Azure Maps Web SDK Samples). I am calling the following function: https://api.nvcf.nvidia.com/v2/nvcf/exec/functions/b0ac1… --- ## [cuOpt for Supply Chain - Optimization scenarios supported](https://forums.developer.nvidia.com/t/cuopt-for-supply-chain-optimization-scenarios-supported/304780)
**Author:** [@sriram10](https://forums.developer.nvidia.com/u/sriram10)\ **Replies:** 3\ **Last updated:** [September 15, 2024, 5:51am UTC](https://forums.developer.nvidia.com/t/cuopt-for-supply-chain-optimization-scenarios-supported/304780 "2024-09-15T05:51:13Z")
Hello Team, We are planning a product focused on Supply Chain planning and modeling. Nvidia cuOpt seems to be a class apart for the Vehicle Routing Problem. We came across the following blog article and the embedded vi… --- ## [cuOpt pricing and deployment options](https://forums.developer.nvidia.com/t/cuopt-pricing-and-deployment-options/305395)
**Author:** [@sriram10](https://forums.developer.nvidia.com/u/sriram10)\ **Replies:** 0\ **Last updated:** [September 2, 2024, 10:56am UTC](https://forums.developer.nvidia.com/t/cuopt-pricing-and-deployment-options/305395 "2024-09-02T10:56:35Z")
Hello Team, One of the many options to run cuOpt is on an Azure VM that runs Nvidia Enterprise AI. ref: Microsoft Azure Marketplace The cost of running Nvidia AI enterprise with a 4 GPU plan in Southeast Asia is around… --- ## [Cuopt more than 3 capacities](https://forums.developer.nvidia.com/t/cuopt-more-than-3-capacities/304241)
**Author:** [@kobe.kwanten](https://forums.developer.nvidia.com/u/kobe.kwanten)\ **Replies:** 0\ **Last updated:** [August 22, 2024, 12:38pm UTC](https://forums.developer.nvidia.com/t/cuopt-more-than-3-capacities/304241 "2024-08-22T12:38:24Z")
Hi there! I’m currently working with cuopt for a proof of concept solution. My client has a lot of task pairs with combinational restrictions. For example, there is a type of tasks x and a type y where tasks in catego… --- ## [Order Vehicle Match error on v24.07](https://forums.developer.nvidia.com/t/order-vehicle-match-error-on-v24-07/301203)
**Author:** [@gordrick](https://forums.developer.nvidia.com/u/gordrick)\ **Replies:** 3\ **Last updated:** [July 29, 2024, 2:39pm UTC](https://forums.developer.nvidia.com/t/order-vehicle-match-error-on-v24-07/301203 "2024-07-29T14:39:28Z")
Hello. Is anyone experiencing the following error in the latest release of cuOpt: When you add a vehicle/order match when only one vehicle is passed it crashes. Multiple vehicles is fine. [Next page](https://forums.developer.nvidia.com/c/ai-data-science/nvidia-cuopt/514.md?page=1)