clopen.solutions
OP-00697 Solution claimed

The modified integer round-up conjecture for cutting stock

Submit a result

Fix a positive integer stock length $W$ and $n \ge 1$ item types; type $i$ has integer length $1 \le \ell_i \le W$ and positive integer demand $b_i$. A cutting pattern is a vector in $$\mathcal{P} = \Bigl\{a \in \mathbb{Z}_{\ge 0}^n : \sum_{i=1}^n \ell_i a_i \le W\Bigr\}.$$ Every capacity-feasible pattern is used: there is no restriction $a_i \le b_i$, so patterns may contain more copies of an item than its demand (nonproper patterns). Let $$z_{\mathrm{IP}} = \min\Bigl\{\sum_{a \in \mathcal{P}} x_a : \sum_{a \in \mathcal{P}} a_i x_a \ge b_i \ (1 \le i \le n),\ x \in \mathbb{Z}_{\ge 0}^{\mathcal{P}}\Bigr\}$$ be the minimum number of stock pieces needed to meet all demands, each use of a pattern counting as one piece, and let $z_{\mathrm{LP}}$ be the value of its linear relaxation, with $x \in \mathbb{R}_{\ge 0}^{\mathcal{P}}$ (the Gilmore–Gomory relaxation). There are instances with $z_{\mathrm{IP}} = \lceil z_{\mathrm{LP}} \rceil + 1$. Karmarkar and Karp's rounding bounds the additive gap $z_{\mathrm{IP}} - z_{\mathrm{LP}}$ by $O(\log^2 z_{\mathrm{IP}})$, and Hoberg and Rothvoss improved the bound to $O(\log z_{\mathrm{LP}})$.

Problem. Does every such instance, with arbitrary item types, lengths, stock length and demands, satisfy $$z_{\mathrm{IP}} \le \lceil z_{\mathrm{LP}} \rceil + 1?$$

Source: Scheithauer and Terno (1995, 1997), the modified integer round-up property (MIRUP). Stated by AIM — Open Applied Problems in the form of Franco da Silva, Loti de Lima, Schouery, Côté and Iori (2026), with nonproper patterns.

Imported from github.com/MColbrook/AIM.

Classification

Primary
OC — Optimization and Control
Secondary
CO
Keywords
AIM Open Applied Problems, bin packing, cutting stock problem, integrality gap

References

@article{scheithauer1995,
  author  = {Scheithauer, Guntram and Terno, Johannes},
  title   = {The modified integer round-up property of the one-dimensional cutting stock problem},
  journal = {European J. Oper. Res.},
  volume  = {84},
  number  = {3},
  year    = {1995},
  pages   = {562--571},
  doi     = {10.1016/0377-2217(95)00022-I},
}
@article{scheithauer1997,
  author  = {Scheithauer, Guntram and Terno, Johannes},
  title   = {Theoretical investigations on the modified integer round-up property for the one-dimensional cutting stock problem},
  journal = {Oper. Res. Lett.},
  volume  = {20},
  number  = {2},
  year    = {1997},
  pages   = {93--100},
  doi     = {10.1016/S0167-6377(96)00047-8},
}
@misc{francodasilva2026,
  author        = {Franco da Silva, Renan Fernando and Loti de Lima, Vin{\'i}cius and Schouery, Rafael C. S. and C{\^o}t{\'e}, Jean-Fran{\c{c}}ois and Iori, Manuel},
  title         = {Polynomial and Pseudopolynomial Algorithms for Two Classes of Bin Packing Instances},
  year          = {2026},
  eprint        = {2604.05152},
  archivePrefix = {arXiv},
}
@inproceedings{hoberg2017,
  author        = {Hoberg, Rebecca and Rothvoss, Thomas},
  title         = {A logarithmic additive integrality gap for bin packing},
  booktitle     = {Proceedings of the Twenty-Eighth Annual ACM-SIAM Symposium on Discrete Algorithms (SODA 2017)},
  publisher     = {SIAM},
  year          = {2017},
  pages         = {2616--2625},
  doi           = {10.1137/1.9781611974782.172},
  eprint        = {1503.08796},
  archivePrefix = {arXiv},
}
@misc{aim-modified-integer-round-up,
  author       = {Brunton, Steve and Colbrook, Matthew J. and de Hoop, Maarten V. and Stepaniants, George and Townsend, Alex and Ward, Rachel},
  title        = {{The modified integer round-up conjecture for cutting stock}},
  howpublished = {AIM --- Open Applied Problems, GitHub repository},
  year         = {2026},
  url          = {https://github.com/MColbrook/AIM/blob/1b7c7fd251c1643bb41f0e3cef38704bcef410fd/problems/341-modified-integer-round-up.md},
  note         = {Entry 341 at commit 1b7c7fd, 2026-10-13},
}

Resolution claims

OpenAI

We disprove the Modified Integer Round-Up Conjecture for bin packing by constructing instances with arbitrarily large additive gaps between the configuration-LP value and the integral optimum. We also prove that, for every fixed nonnegative integer $c$, distinguishing instances that fit in $B$ bins from those requiring more than $B+c$ bins is NP-hard. Both results hold with rational item sizes greater …

Counterexample · not verified · fidelity not established · posted on GitHub 6 Oct 2026

Status history

WhenChangeByReason
2026-10-16 Needs verification → Solution claimed — Recorded here: math @ 1581d4c, source/main.tex.

Citing this problem

This identifier is permanent. It will not change if the problem is solved, reclassified or withdrawn.

OP-00697, "The modified integer round-up conjecture for cutting stock", clopen.solutions.