B$^3$-PWL: GPU-Batched Branch-and-Bound for Piecewise-Linear Optimization with SOS2 Constraints

Yilin Guan; Shuqing Luo; Pingzhi Li; Tianlong Chen; Kaidi Xu

B$^3$-PWL: GPU-Batched Branch-and-Bound for Piecewise-Linear Optimization with SOS2 Constraints

Yilin Guan, Shuqing Luo, Pingzhi Li, Tianlong Chen, Kaidi Xu

arXivarxiv/2608.28988Aug 29, 2026math.OCscore 0

Piecewise-linear (PWL) optimization problems arise in many mixed-integer programming (MIP) optimization applications, including portfolio optimization, workforce scheduling, and resource allocation. But solving them to global optimality remains computationally expensive because branch-and-bound repeatedly solves LP relaxation subproblems. Existing solvers are largely CPU-centric, leaving the scalability of modern GPUs underutilized. Few prior GPU-accelerated branch-and-bound either targets neural network which is not suitable for general PWL optimization, or accelerates only auxiliary…

Source JSON

Comments and reviews

No comments yet.

API

curl /api/v1/records/arxiv/2608.28988
curl -X POST /api/v1/records/arxiv/2608.28988/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2608.28988/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'