Predicting Runtime and Resource Utilization of Jobs on Integrated Cloud and HPC Systems

This is a time series data set of resource utilizations and runtime for jobs run on both HPC systems (IC2 at Brookhaven Naional Lab, Polaris at Argonne National Lab and Amazon Web Services. The data set can be used for machine learning models to predict runtime and resource utilization of jobs on a variety of systems.本数据集为时间序列数据集，涵盖了运行于两类高性能计算（High Performance Computing, HPC）系统及亚马逊云服务（Amazon Web Services）上的作业的资源利用率与运行时长数据，其中两类高性能计算系统分别为布鲁克海文国家实验室的IC2与阿贡国家实验室的Polaris。本数据集可用于训练机器学习模型，以预测各类计算平台上作业的运行时长与资源利用率。