精准快速获取计算机系统的实时功耗是功耗优化研究的基础,因此提出并建立了一种高精度的计算机功耗估算模型。通过分析统计系统运行时代表性的性能计数事件,应用机器学习理论分析性能事件与功耗的关系,建立多核计算机系统实时功耗估算...精准快速获取计算机系统的实时功耗是功耗优化研究的基础,因此提出并建立了一种高精度的计算机功耗估算模型。通过分析统计系统运行时代表性的性能计数事件,应用机器学习理论分析性能事件与功耗的关系,建立多核计算机系统实时功耗估算模型。模型构建时使用多元线性回归(multiple linear regression,MLR)方法以及支持向量回归(support vector regression,SVR)方法分析两者关系,并对两种方法建立的功耗估算模型进行了对比分析。实验结果表明,基于性能事件的功耗估算模型可准确估计计算机实时功耗,估算误差不高于3%。与已有模型相比较,该估算模型精度更高、通用性更好。展开更多
This paper discusses the analysis done on the meteorological ocean buoy mooring used for monitoring the Indian seas. Based on the extreme environmental parameters experienced by the buoys, mooring loads are analyzed u...This paper discusses the analysis done on the meteorological ocean buoy mooring used for monitoring the Indian seas. Based on the extreme environmental parameters experienced by the buoys, mooring loads are analyzed using offshore dynamic analysis software. The results obtained are validated with the tension recorder installed in one of the moorings, and the results are found to comply with an accuracy of better than 1%. The successful on demand performance of the mooring during major cyclones in the Bay of Bengal and the vital meteorological and oceanographic information provided by the buoy during these disastrous cyclonic events validates the mooring design, and proves the data availability for societal needs. The time critical data assimilated in the cyclone prediction models have given confidence to improve the country's weather prediction and climate modelling capabilities.展开更多
文摘精准快速获取计算机系统的实时功耗是功耗优化研究的基础,因此提出并建立了一种高精度的计算机功耗估算模型。通过分析统计系统运行时代表性的性能计数事件,应用机器学习理论分析性能事件与功耗的关系,建立多核计算机系统实时功耗估算模型。模型构建时使用多元线性回归(multiple linear regression,MLR)方法以及支持向量回归(support vector regression,SVR)方法分析两者关系,并对两种方法建立的功耗估算模型进行了对比分析。实验结果表明,基于性能事件的功耗估算模型可准确估计计算机实时功耗,估算误差不高于3%。与已有模型相比较,该估算模型精度更高、通用性更好。
文摘This paper discusses the analysis done on the meteorological ocean buoy mooring used for monitoring the Indian seas. Based on the extreme environmental parameters experienced by the buoys, mooring loads are analyzed using offshore dynamic analysis software. The results obtained are validated with the tension recorder installed in one of the moorings, and the results are found to comply with an accuracy of better than 1%. The successful on demand performance of the mooring during major cyclones in the Bay of Bengal and the vital meteorological and oceanographic information provided by the buoy during these disastrous cyclonic events validates the mooring design, and proves the data availability for societal needs. The time critical data assimilated in the cyclone prediction models have given confidence to improve the country's weather prediction and climate modelling capabilities.