发表机构
Universidad Industrial de Santander; INSA Lyon; INRIA; Université Grenoble-Alpes; Grenoble-INP(桑坦德工业大学; 里昂国立应用科学学院; 法国国家信息与自动化研究所; 格勒诺布尔阿尔卑斯大学; 格勒诺布尔国立理工学院)
机构由 AI 辅助整理,请以论文原文为准。AI 中文总结
本文提出多维度量框架,结合性能、利用率与可持续性指标,在混合测试平台上揭示权衡关系,指导HPC系统部署与调度优化,兼顾高性能与可持续性。
AI 中文摘要
从传统高性能计算(HPC)向计算连续体的转变,强调了跨多尺度混合架构的高效资源管理和可持续实践。本文引入了一个多维度量框架,用于表征这些系统并指导现代工作负载的部署策略。该框架结合了架构性能指标(如吞吐量、延迟、可扩展性)、系统利用率,以及关键的可持续性和准确性指标(如能效和功耗)。利用模块化混合测试平台,实验揭示了指标之间复杂的关系,尤其是准确性与能耗之间的权衡,以及混合节点的效率。这些指南有助于识别最佳运行点,并为改进编排器和调度器(如Kubernetes)奠定基础,以将要求苛刻的应用(包括人工智能和量子计算)分配到合适的系统模块,确保高性能和可持续性。
英文摘要
The transition from traditional High Performance Computing (HPC) to the Computing Continuum emphasizes efficient resource management and sustainable practices across Multi-Scale hybrid architectures. This paper introduces a multidimensional metric framework to characterize these systems and guide deployment strategies for modern workloads. The framework combines Architectural Performance metrics (such as Throughput, Latency, Scalability), System Utilization, and key Sustainability and Accuracy indicators (such as Energy Efficiency and Power Consumption). Using a modular hybrid testbed, experiments reveal complex relationships among metrics, especially the trade-offs between accuracy and energy, and the efficiency of hybrid nodes. The guidelines help identify optimal operating points and lay the groundwork for improving orchestrators and schedulers (e.g., Kubernetes) to assign demanding applications, including AI and Quantum Computing, to suitable system modules, ensuring high performance and sustainability.