analiza aiqbench
The paper applies machine learning to a practical infrastructure problem: estimating a virtual server's CPU-package power from metrics visible inside the VM, using host-side RAPL readings as training targets. Its useful finding is the limit on generalization. Predictions are reliable when deployment conditions appear in training, but zero-shot transfer to unseen workloads, hardware, or configurations is weaker, so anyone adopting this for energy accounting needs representative training data for their own fleet.
Energy Estimation for Virtual Servers Using Resource Utilization and Machine Learning
arXiv:2509.09991v2 Announce Type: replace Abstract: Estimating the energy use of virtual servers is difficult because physical power interfaces and privileged host counters are typically accessible only on the underlying physical host. We investigate a machine learning approach that predicts the CPU-package power attributable to an individual virtual server from system and performance metrics observable within that virtual server itself. During training, host-side RAPL measurements provide the power target used to learn this mapping, while deployed inference requires only measurements available within the virtual server. We evaluate the approach across heterogeneous CPU platforms, multiple workload families, unseen workloads and hardware, concurrent workload combinations, and varying resource allocations using temporally blocked evaluation. Prediction is generally strong when deployment conditions are represented during training, while zero-shot transfer to unseen workloads, hardware, and configurations is less reliable. A pooled multi-platform model retains good performance when target platforms are represented during training, and pooled multi-configuration modeling performs well across different processor and memory allocations. Feature-importance analysis reveals interpretable resource signatures consistent with workload behavior across CPU, memory, storage, and network activity. Monitoring and inference add little direct CPU-power overhead. The results support practical virtual server power estimation when training data cover the intended deployment domain.
Abstract:Estimating the energy use of virtual servers is difficult because physical power interfaces and privileged host counters are typically accessible only on the underlying physical host. We investigate a machine learning approach that predicts the CPU-package power attributable to an individual virtual server from system and performance metrics observable within that virtual server itself. During training, host-side RAPL measurements provide the power target used to learn this mapping, while deployed inference requires only measurements available within the virtual server. We evaluate the approach across heterogeneous CPU platforms, multiple workload families, unseen workloads and hardware, concurrent workload combinations, and varying resource allocations using temporally blocked evaluation. Prediction is generally strong when deployment conditions are represented during training, while zero-shot transfer to unseen workloads, hardware, and configurations is less reliable. A pooled multi-platform model retains good performance when target platforms are represented during training, and pooled multi-configuration modeling performs well across different processor and memory allocations. Feature-importance analysis reveals interpretable resource signatures consistent with workload behavior across CPU, memory, storage, and network activity. Monitoring and inference add little direct CPU-power overhead. The results support practical virtual server power estimation when training data cover the intended deployment domain.
Machine Learning (cs.LG)
arXiv:2509.09991 [cs.LG]
(or arXiv:2509.09991v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2509.09991 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 287 cuvinte
Subiecte conexe
Dense Process Supervision for Search Agents via Fact Utility Estimation
arXiv · 2 sept. 2026
SkillScaffold: Scaffolding Agentic Reinforcement Learning with Policy-Adaptive Skill Credit Assignment and Selective Injection
arXiv · 14 oct. 2026
How Jailbreak Attacks Inform Safety Alignment: A Defender-Centric, Shapley-Based Evaluation of Jailbreak Contributions
arXiv · 21 iul. 2026
RM-WAM: Reliable Memory-Augmented World Action Models for Long-Horizon Robot Manipulation
arXiv · 14 oct. 2026