Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.00987 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911105745420288 |
|---|---|
| author | Zhang, Zhiwei Shen, Jiayu Kumar, Niraj Pistoia, Marco |
| author_facet | Zhang, Zhiwei Shen, Jiayu Kumar, Niraj Pistoia, Marco |
| contents | Low Autocorrelation Binary Sequences (LABS) is a particularly challenging binary optimization problem which quickly becomes intractable in finding the global optimum for problem sizes beyond 66. This aspect makes LABS appealing to use as a test-bed for meta-heuristic optimization solvers to target large problem sizes. In this work, we introduce a massively parallelized implementation of the memetic tabu search algorithm to tackle LABS problem for sizes up to 120. By effectively combining the block level and thread level parallelism framework within a single Nvidia-A100 GPU, and creating hyper optimized binary-valued data structures for shared memory among the blocks, we showcase up to 26 fold speedup compared to the analogous 16-core CPU implementation. Our implementation has also enabled us to find new LABS merit factor values for sixteen different problem sizes between 92 and 120. Crucially, we also showcase improved values for five odd-sized problems {99, 107, 109, 113, 119} whose previous best known results coincided with the provably optimal skew-symmetric search sequences. Consequently, our result highlights the importance of a focus on general-purpose solver to tackle LABS, since leveraging its skew-symmetry could lead to sub-optimal solutions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_00987 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | New Improvements in Solving Large LABS Instances Using Massively Parallelizable Memetic Tabu Search Zhang, Zhiwei Shen, Jiayu Kumar, Niraj Pistoia, Marco Distributed, Parallel, and Cluster Computing Low Autocorrelation Binary Sequences (LABS) is a particularly challenging binary optimization problem which quickly becomes intractable in finding the global optimum for problem sizes beyond 66. This aspect makes LABS appealing to use as a test-bed for meta-heuristic optimization solvers to target large problem sizes. In this work, we introduce a massively parallelized implementation of the memetic tabu search algorithm to tackle LABS problem for sizes up to 120. By effectively combining the block level and thread level parallelism framework within a single Nvidia-A100 GPU, and creating hyper optimized binary-valued data structures for shared memory among the blocks, we showcase up to 26 fold speedup compared to the analogous 16-core CPU implementation. Our implementation has also enabled us to find new LABS merit factor values for sixteen different problem sizes between 92 and 120. Crucially, we also showcase improved values for five odd-sized problems {99, 107, 109, 113, 119} whose previous best known results coincided with the provably optimal skew-symmetric search sequences. Consequently, our result highlights the importance of a focus on general-purpose solver to tackle LABS, since leveraging its skew-symmetry could lead to sub-optimal solutions. |
| title | New Improvements in Solving Large LABS Instances Using Massively Parallelizable Memetic Tabu Search |
| topic | Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2504.00987 |