Skip to content

Commit 6d11a00

Browse files
authored
Release notes for v25.10 (#432)
## Issue Authors: - Cindy Wilkinson (https://github.com/cwilkinson76) - Ramakrishnap (https://github.com/rgsl888prabhu) - Chris Maes (https://github.com/chris-maes) Approvers: - Trevor McKay (https://github.com/tmckayus) URL: #432
1 parent 91a19f8 commit 6d11a00

File tree

1 file changed

+87
-0
lines changed

1 file changed

+87
-0
lines changed

docs/cuopt/source/release-notes.rst

Lines changed: 87 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2,6 +2,93 @@
22
Release Notes
33
=====================
44

5+
====================
6+
Release Notes 25.10
7+
====================
8+
9+
New Features (25.10)
10+
--------------------
11+
12+
- New barrier method for solving LPs. Uses cuDSS for sparse Cholesky / LDT.
13+
- Concurrent mode for LPs now uses PDLP, dual simplex, and barrier
14+
- New PDLP solver mode Stable3.
15+
- MIP presolve using Papilo (enabled by default). LP presolve using Papilo (optional).
16+
- Parallel branch and bound on the CPU: multiple best-first search and diving threads
17+
18+
Breaking Changes (25.10)
19+
------------------------
20+
21+
- New PDLP Solver mode Stable3 is the default
22+
23+
24+
Improvements (25.10)
25+
--------------------
26+
27+
- Add setting "CUOPT_BARRIER_DUAL_INITIAL_POINT" to change the dual initial point used by barrier
28+
- CPUFJ for local search + simple rounding
29+
- FP as a local search
30+
- Sub-MIP recombiner and B&B global variable changes
31+
- Implement GF(2) presolve reduction
32+
- Implement node presolve
33+
- CUDA 13/12.9 support
34+
- Build and test with CUDA 13.0.0
35+
- Add read/write MPS and relaxation to python API
36+
- Decompression for ``.mps.gz`` and ``.mps.bz2`` files
37+
- Enable parallelism for root node presolve
38+
- Enable singleton stuffing and use Papilo default params
39+
- Make infeasibility checks consistent between the main solver and presolver
40+
- Add maximization support for root node presolve
41+
- Performance improvement in dual simplex's right-looking LU factorization
42+
- Fix high GPU memory usage
43+
- Print cuOpt version / machine info before solving
44+
- ``cuopt-server``: update dependencies (drop httpx, add psutil)
45+
- Add nightly testing of cuOpt jump interface
46+
- Compression tests are not run when compression is disabled
47+
- Add sanitizer build option- Heuristic Improvements: balance between generation and improvement heuristics
48+
- Loosen presolve tolerance and update timers to report cumulative presolve/solve time
49+
- Warn in case a dependent library is not found in libcuopt load
50+
- Combined variable bounds
51+
- Add Commit Sha to container for reference
52+
- use GCC 14, consolidate dependency groups, update pre-commit hooks
53+
- Add support for nightly ``cuopt-examples`` notebook testing
54+
- Reduce hard-coded version usage in repo
55+
- Container to work on all different users including root
56+
- Changes to download LP and MILP datasets, and also disable cvxpy testing for 3.10
57+
- Faster engine compile time
58+
- Fix pre-commit for trailing whitespace and end of file
59+
- Merge update version and fix version format bugs
60+
- This library now supports the QPS format, which is an extension of the standard MPS format for representing quadratic programming problems.
61+
62+
63+
Bug Fixes (25.10)
64+
-----------------
65+
66+
- Fix variables out of bounds caused by CPUFJ LP scratch thread
67+
- Fix the maybe-uninitialized compilation error
68+
- Fix linking errors in the test suite when disabling C adaptor
69+
- Compute relative gap with respect to user objectives
70+
- Add http timeout values for general, send, and receive to client
71+
- Fix bug in ``fixed_problem_computation``
72+
- Remove ``limiting_resource_adaptor`` leftover
73+
- Add support for cuda13 container and fix cuda13 lib issues in wheel
74+
- Return Infeasible if the user problem contains crossing bounds
75+
- Fix out-of-bound access in ``clean_up_infeasibilities``
76+
- Empty columns with infinite bounds are not removed
77+
78+
79+
Documentation (25.10)
80+
---------------------
81+
82+
- Add tutorial video links to Decompression
83+
- Add warmstart, model update, update docs
84+
- add docs on CI workflow inputs
85+
- Add name to drop-down for video link
86+
- Add video link to the docs and to the Readme
87+
- Add documentation on nightly installation commands
88+
- Fix version in version tab, change log, and fix typos
89+
- Doc update for container version update, and add ``nvidia-cuda-runtime`` as a dependency
90+
91+
592
====================
693
Release Notes 25.08
794
====================

0 commit comments

Comments
 (0)