SLA-MORL: SLA-Aware Multi-Objective Reinforcement Learning for HPC Resource Optimization

arXiv 2508.03509 ↗First posted 5 Aug 2025cs.LGcited by 0

On the map: Deep reinforcement learning methods → RL for energy, networks and operations

Abstract (arXiv)

Dynamic resource allocation for machine learning workloads in cloud environments remains challenging due to competing objectives of minimizing training time and operational costs while meeting Service Level Agreement (SLA) constraints. Traditional approaches employ static resource allocation or single-objective optimization, leading to either SLA violations or resource waste. We present SLA-MORL, an adaptive multi-objective reinforcement learning framework that intelligently allocates GPU and CPU resources based on user-defined preferences (time, cost, or balanced) while ensuring SLA compliance. Our approach introduces two key innovations: (1) intelligent initialization through historical learning or efficient baseline runs that eliminates cold-start problems, reducing initial exploration overhead by 60%, and (2) dynamic weight adaptation that automatically adjusts optimization priorities based on real-time SLA violation severity, creating a self-correcting system. SLA-MORL constructs a 21-dimensional state representation capturing resource utilization, training progress, and SLA compliance, enabling an actor-critic network to make informed allocation decisions across 9 possible actions. Extensive evaluation on 13 diverse ML workloads using production HPC infrastructure demonstrates that SLA-MORL achieves 67.2% reduction in training time for deadline-critical jobs, 68.8% reduction in costs for budget-constrained workloads, and 73.4% improvement in overall SLA compliance compared to static baselines. By addressing both cold-start inefficiency and dynamic adaptation challenges, SLA-MORL provides a practical solution for cloud resource management that balances performance, cost, and reliability in modern ML training environments.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for energy, networks and operations, one of 15 directions in Deep reinforcement learning methods. Builds on 12 of its 30 references that are in the map. First posted on arXiv 5 Aug 2025.

Builds on

12 of 30 references.

Closest ancestors · 1

Same direction, newest first

2025Deep Reinforcement Learning for Job Scheduling and Resource Management in Cloud Computing: An Algorithm-Level Review2 Jan

Other references · 11

2023Efficient Online Reinforcement Learning with Offline Datacited by 363
2022When to Trust Your Simulator: Dynamics-Aware Hybrid Offline-and-Online Reinforcement Learningcited by 60
2023Adaptive Policy Learning for Offline-to-Online Reinforcement Learningcited by 40
2023Train Once, Get a Family: State-Adaptive Balances for Offline-to-Online Reinforcement Learningcited by 26
2025Pareto Set Learning for Multi-Objective Reinforcement Learningcited by 20
2017A Hierarchical Framework of Cloud Resource Allocation and Power Management Using Deep Reinforcement Learningcited by 16
2024Offline Multi-Agent Reinforcement Learning via In-Sample Sequential Policy Optimizationcited by 15
2024gWaveNet: Classification of Gravity Waves from Noisy Satellite Data using Custom Kernel Integrated Deep Learning Methodcited by 5
2024A HPC Co-Scheduler with Reinforcement Learningcited by 2
2025Joint Retrieval of Cloud properties using Attention-based Deep Learning Modelscited by 2
2020MARS: Malleable Actor-Critic Reinforcement Learning Schedulercited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way