【摘要】 异构调度可使大规模计算系统采用并行方式聚合广域分布的各种资源以提高性能.传统调度目标追求高性能而忽视高效能,远不能适应绿色计算科学发展新要求.因此,在理论上,一方面基于对动态频率和电压等系统参数的精细表述及有效量化,建立面向协同异构计算且易于复用的能效感知云调度模型;另一方面,提出并实现适于超计算机混合体系的多学科背景的元启发式多目标全局优化算法.从技术上解决了面向不同环境目标的云调度实施条件界定及其调度指标(时间、能效)实时变化描述等问题.大量仿真实验结果表明:与3个元启发式云调度器相比,该方法在能效及可扩展等方面优势明显;对于高维实例,整体性能改善分别达到8%,12%和14%.更多还原

【Abstract】 Designed to provide pervasive access to distributed resources in parallel ways,heterogeneous scheduling is extensively applied in large-scale computing system for its high performance.Conventional real-time scheduling algorithms,however,often overlook energy-efficiency while focusing on stringent timing constraints.To engage in green heterogeneous computing,a reusable energy-aware cloud model is first presented via mathematical formulation and quantization of the system parameters such as dynamic voltage and frequency scaling(DVFS),and dynamic power management(DPM).In addition,multidisciplinary context for multi-objective global optimization meta-heuristic is proposed and accomplished based on the supercomputer hybrid architecture.Furthermore,some technological breakthroughs are achieved with respect to boundary conditions for different heterogeneous computing and cloud scheduling,and descriptions of real-time variation of scheduling indexes(stringent timing constraints and energy-efficiency).Extensive simulation experiments highlight higher efficacy and better scalability for the proposed approaches compared with the other three meta-heuristics; the overall improvements achieve 8%,12% and 14% for high-dimension instances.更多还原