Beyond a Single Queue: Multi-Level-Multi-Queue as an Effective Design for SSSP problems on GPUs
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910018087944192 |
|---|---|
| author | Hu, Zhengding Sun, Jingwen Jiang, Le Wang, Yuhao Lin, Junqing Zong, Yi Sun, Guangzhong |
| author_facet | Hu, Zhengding Sun, Jingwen Jiang, Le Wang, Yuhao Lin, Junqing Zong, Yi Sun, Guangzhong |
| contents | As one of the most fundamental problems in graph processing, the Single-Source Shortest Path (SSSP) problem plays a critical role in numerous application scenarios. However, existing GPU-based solutions remain inefficient, as they typically rely on a single, fixed queue design that incurs severe synchronization overhead, high memory latency, and poor adaptivity to diverse inputs. To address these inefficiencies, we propose MultiLevelMultiQueue (MLMQ), a novel data structure that distributes multiple queues across the GPU's multi-level parallelism and memory hierarchy. To realize MLMQ, we introduce a cache-like collaboration mechanism for efficient inter-queue coordination, and develop a modular queue design based on unified Read and Write primitives. Within this framework, we expand the optimization space by designing a set of GPU-friendly queues, composing them across multiple levels, and further providing an input-adaptive MLMQ configuration scheme. Our MLMQ design achieves average speedups of 1.87x to 17.13x over state-of-the-art implementations. Our code is open-sourced at https://github.com/Leo9660/MLMQ.git. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_10080 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Beyond a Single Queue: Multi-Level-Multi-Queue as an Effective Design for SSSP problems on GPUs Hu, Zhengding Sun, Jingwen Jiang, Le Wang, Yuhao Lin, Junqing Zong, Yi Sun, Guangzhong Data Structures and Algorithms As one of the most fundamental problems in graph processing, the Single-Source Shortest Path (SSSP) problem plays a critical role in numerous application scenarios. However, existing GPU-based solutions remain inefficient, as they typically rely on a single, fixed queue design that incurs severe synchronization overhead, high memory latency, and poor adaptivity to diverse inputs. To address these inefficiencies, we propose MultiLevelMultiQueue (MLMQ), a novel data structure that distributes multiple queues across the GPU's multi-level parallelism and memory hierarchy. To realize MLMQ, we introduce a cache-like collaboration mechanism for efficient inter-queue coordination, and develop a modular queue design based on unified Read and Write primitives. Within this framework, we expand the optimization space by designing a set of GPU-friendly queues, composing them across multiple levels, and further providing an input-adaptive MLMQ configuration scheme. Our MLMQ design achieves average speedups of 1.87x to 17.13x over state-of-the-art implementations. Our code is open-sourced at https://github.com/Leo9660/MLMQ.git. |
| title | Beyond a Single Queue: Multi-Level-Multi-Queue as an Effective Design for SSSP problems on GPUs |
| topic | Data Structures and Algorithms |
| url | https://arxiv.org/abs/2602.10080 |