NVIDIA/cuopt v26.06.00

v26.06.00

Repository: NVIDIA/cuopt

Tag: v26.06.00

Published: 2026-06-09T22:05:51Z

Prerelease: no

Release notes:

What's Changed

🚨 Breaking Changes

• ci(pr): gate builds on changed-files and broaden safe exclusions by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1118
• Migrate RMM usage to CCCL MR design by @bdice in https://github.com/NVIDIA/cuopt/pull/1035
• Fold libmps_parser into libcuopt by @mlubin in https://github.com/NVIDIA/cuopt/pull/1193

🐛 Bug Fixes

• Add missing thrust headers across C++ sources by @bdice in https://github.com/NVIDIA/cuopt/pull/997
• Update sonarcube host link to Nvidia by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1015
• Require openssl so that conda verson is found before system lib by @hlinsen in https://github.com/NVIDIA/cuopt/pull/1084
• fix(cuda_helpers): clear sticky error and avoid cache poisoning in set_shmem_of_kernel by @np96 in https://github.com/NVIDIA/cuopt/pull/1095
• Fix double va_start undefined behavior in error handling by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1086
• Fixed lower bound in single thread mode by @nguidotti in https://github.com/NVIDIA/cuopt/pull/1111
• test(c_api): skip flaky TimeLimitTestFixture MIP case (#1135) by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1136
• Replace flaky determinism test by @aliceb-nv in https://github.com/NVIDIA/cuopt/pull/1141
• Use fp64 for all node stack buffers and fix indexing in vehicle order match test by @hlinsen in https://github.com/NVIDIA/cuopt/pull/1137
• fix(benchmarks): drop removed rmm owning_wrapper; build in CI by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1143
• Fix loose tolerances in the objective function integrality detection by @aliceb-nv in https://github.com/NVIDIA/cuopt/pull/1148
• ci(docker): pin all first-party cuOpt wheels to CUOPT_VER by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1157
• Fix CI crash on RTX6000 by @nguidotti in https://github.com/NVIDIA/cuopt/pull/1184
• miplib_test: allow tiny error on objective equality by @mlubin in https://github.com/NVIDIA/cuopt/pull/1199
• Fix concurrent LP exception cleanup by @mlubin in https://github.com/NVIDIA/cuopt/pull/1206
• Allow nested parallel regions for determinism mode by @nguidotti in https://github.com/NVIDIA/cuopt/pull/1201
• Fix destruction order in solve.cu by @mlubin in https://github.com/NVIDIA/cuopt/pull/1216
• Fix by-reference capture of local variable in solve.cu by @mlubin in https://github.com/NVIDIA/cuopt/pull/1214
• Fix route priority sort indexing by @bdice in https://github.com/NVIDIA/cuopt/pull/1222
• Drop constness in raft handle for data_model_view_t (routing) by @nguidotti in https://github.com/NVIDIA/cuopt/pull/1241
• Fix lost nodes in branch and bound by @hlinsen in https://github.com/NVIDIA/cuopt/pull/1138
• Fix vehicle fixed cost accounting in fragment-vs-route deltas by @mlubin in https://github.com/NVIDIA/cuopt/pull/1244
• Fix an out of bounds access error by @rg20 in https://github.com/NVIDIA/cuopt/pull/1245
• Fix a bug in dual variables and reduced costs when we add implied bounds on free variables by @chris-maes in https://github.com/NVIDIA/cuopt/pull/1237
• Cuda graph replays on capture error by @akifcorduk in https://github.com/NVIDIA/cuopt/pull/1253
• Fix accidentally quadratic add_slacks_to_basis in crossover by @nguidotti in https://github.com/NVIDIA/cuopt/pull/1268
• Replace __constant__ FJ settings with per-instance device_scalar by @mlubin in https://github.com/NVIDIA/cuopt/pull/1284
• Stabilize logger config in LP BatchSolve and version logging by @mlubin in https://github.com/NVIDIA/cuopt/pull/1286
• Add extra numerical guards for flow cover cuts by @aliceb-nv in https://github.com/NVIDIA/cuopt/pull/1316
• Fix broken sample.mps download link in server LP examples docs by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1358
• Fix for maximization problems with quadratic constraints by @chris-maes in https://github.com/NVIDIA/cuopt/pull/1377
• Use 26.6 (not 26.06) in README pip install commands by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1380

📖 Documentation

• Tweak testing instructions in CONTRIBUTING.md by @mlubin in https://github.com/NVIDIA/cuopt/pull/1000
• Exclude L1TESTs from contributing instructions by @mlubin in https://github.com/NVIDIA/cuopt/pull/1034
• CONTRIBUTING.md: Recommend local conda environments by @mlubin in https://github.com/NVIDIA/cuopt/pull/1085
• docs(readme): use py3.13 docker tags (py3.14 not supported) by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1112
• Fix typo in Constraint docstring by @mlubin in https://github.com/NVIDIA/cuopt/pull/1113
• update mip scaling doc by @Iroy30 in https://github.com/NVIDIA/cuopt/pull/1121
• docs: migrate support link to GitHub Discussions; tidy doc links by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1144
• Add VRP dimension developer skill by @rg20 in https://github.com/NVIDIA/cuopt/pull/1233
• Fix for cuDSS 0.8; remove incorrect presolve docs in cuopt_cli by @chris-maes in https://github.com/NVIDIA/cuopt/pull/1353
• Update LP/QP/MILP docs to cover 26.06 API additions by @rg20 in https://github.com/NVIDIA/cuopt/pull/1339
• docs: add 26.06 release notes and updates to docs by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1294
• Mark MILP as beta; add QCQP and SOCP to README by @rgsl888prabhu in https://github.com/NVIDIA/cuopt/pull/1384

🚀 New Features

• Add semi continuous variables support by @hlinsen in https://github.com/NVIDIA/cuopt/pull/1096
• Extend batch pdlp support by @Kh4ster in https://github.com/NVIDIA/cuopt/pull/1152
• Implement python and C api for semi-continuous variables by @hlinsen in https://github.com/NVIDIA/cuopt/pull/1225
• Add LP format reader; accept .lp wherever .mps is accepted by @mlubin in https://github.com/NVIDIA/cuopt/pull/1120
• Detect MIP symmetry using dejavu; Perform orbital fixing and lexical reduction by @chris-maes in https://github.com/NVIDIA/cuopt/pull/1103
• Implement C APIs for adding quadratic objectives and constraints by @rg20 in https://github.com/NVIDIA/cuopt/pull/1247
• Add flow cover cuts at root by @hlinsen in https://github.com/NVIDIA/cuopt/pull/1178
• Extend the barrier solver for Second-Order Cone Programming (SOCP) by @chris-maes in https://github.com/NVIDIA/cuopt/pull/1290
• add QCQP support for grpc by @tmckayus in https://github.com/NVIDIA/cuopt/pull/1320

🛠️ Improvements

• Main merge…