Hybrid Learning and Optimization-Based Dynamic Scheduling for DL Workloads on Heterogeneous GPU Clusters
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918243314171904 |
|---|---|
| author | Dongare, Shruti Khan, Redwan Ibne Seraj Albahar, Hadeel Zhao, Nannan Maita, Diego Melendez Butt, Ali R. |
| author_facet | Dongare, Shruti Khan, Redwan Ibne Seraj Albahar, Hadeel Zhao, Nannan Maita, Diego Melendez Butt, Ali R. |
| contents | Modern cloud platforms increasingly host large-scale deep learning (DL) workloads, demanding high-throughput, low-latency GPU scheduling. However, the growing heterogeneity of GPU clusters and limited visibility into application characteristics pose major challenges for existing schedulers, which often rely on offline profiling or application-specific assumptions. We present RLTune, an application-agnostic reinforcement learning (RL)-based scheduling framework that dynamically prioritizes and allocates DL jobs on heterogeneous GPU clusters. RLTune integrates RL-driven prioritization with MILP-based job-to-node mapping to optimize system-wide objectives such as job completion time (JCT), queueing delay, and resource utilization. Trained on large-scale production traces from Microsoft Philly, Helios, and Alibaba, RLTune improves GPU utilization by up to 20%, reduces queueing delay by up to 81%, and shortens JCT by as much as 70 percent. Unlike prior approaches, RLTune generalizes across diverse workloads without requiring per-job profiling, making it practical for cloud providers to deploy at scale for more efficient, fair, and sustainable DL workload management. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_10271 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Hybrid Learning and Optimization-Based Dynamic Scheduling for DL Workloads on Heterogeneous GPU Clusters Dongare, Shruti Khan, Redwan Ibne Seraj Albahar, Hadeel Zhao, Nannan Maita, Diego Melendez Butt, Ali R. Distributed, Parallel, and Cluster Computing Artificial Intelligence Machine Learning Modern cloud platforms increasingly host large-scale deep learning (DL) workloads, demanding high-throughput, low-latency GPU scheduling. However, the growing heterogeneity of GPU clusters and limited visibility into application characteristics pose major challenges for existing schedulers, which often rely on offline profiling or application-specific assumptions. We present RLTune, an application-agnostic reinforcement learning (RL)-based scheduling framework that dynamically prioritizes and allocates DL jobs on heterogeneous GPU clusters. RLTune integrates RL-driven prioritization with MILP-based job-to-node mapping to optimize system-wide objectives such as job completion time (JCT), queueing delay, and resource utilization. Trained on large-scale production traces from Microsoft Philly, Helios, and Alibaba, RLTune improves GPU utilization by up to 20%, reduces queueing delay by up to 81%, and shortens JCT by as much as 70 percent. Unlike prior approaches, RLTune generalizes across diverse workloads without requiring per-job profiling, making it practical for cloud providers to deploy at scale for more efficient, fair, and sustainable DL workload management. |
| title | Hybrid Learning and Optimization-Based Dynamic Scheduling for DL Workloads on Heterogeneous GPU Clusters |
| topic | Distributed, Parallel, and Cluster Computing Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2512.10271 |