CODES: Enabling Co-Design of Multi-Layer Exascale Storage Architectures
CODES: Enabling Co-Design of Multi-Layer Exascale Storage Architectures
复制标题
代码：实现多层百亿亿次存储架构的协同设计
DOI：
--
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
R. Ross
中科院分区：
文献类型：
--
作者：
Jason Cope;Ning Liu;S. Lang;P. Carns;C. Carothers;R. Ross
对于HPC系统设计人员而言，Exascale存储系统的性能和可靠性设计约束是重大挑战。我们正在开发代码模拟工具包，以为系统设计人员提供模拟工具，以便他们更好地了解Exascale存储系统的功能和设计约束。代码的目标是通过为Exascale存储提供详细，准确且高度平行的模拟工具包，启用Exascale存储系统的探索和共同设计。在本文中，我们介绍了允许系统设计人员评估Exascale存储系统设计的代码工具的功能。我们证明了使用代码来评估潜在的Exascale存储网络模型和存储系统功能。
Performance and reliability design constraints for exascale storage systems are significant challenges for HPC system designers. We are developing the CODES simulation toolkit to equip system designers with simulation tools so that they better understand the features and design constraints of exascale storage systems. The goal for CODES is to enable the exploration and co-design of exascale storage systems by providing a detailed, accurate, and highly parallel simulation toolkit for exascale storage. In this paper, we present the capabilities of the CODES tools that allow systems designers to assess exascale storage system designs. We demonstrate the use of CODES to evaluate a potential exascale storage network model and storage system features.