摘要:针对智算仿真难以满足广域网时空动态性需求的情况,提出了一种面向多算力中心协同的广域智算网络仿真架构。该架构的主要创新点包括:提出基于属性图模型的拓扑抽象方法,支持异构算力间的不规则连接建模和不稳定网络还原;提出基于“流感知框架”的广域通信模拟架构,为仿真提供高精度的广域网络通信模拟;提出多算力中心间事件触发的动态调度协议,基于逻辑时钟实现跨域操作因果一致性。本架构的提出弥补了广域多算力中心背景下仿真工具的缺失,为广域智算领域的相关研究人员提供高效、可靠的仿真支持。
关键词:多算力中心协同;广域环境;算网融合;仿真架构
Abstract: In response to the situation that intelligent computing simulation is difficult to meet the requirements of the spatio-temporal dynamics of the wide area network, a wide-area intelligent computing network simulation architecture oriented to the collaboration of multiple computing power centers is proposed. The main innovations of this architecture include: proposing a topology abstraction method based on the property graph model, which supports the modeling of irregular connections between heterogeneous computing powers and the restoration of unstable networks; proposing a wide-area communication simulation architecture based on the "flow awareness framework" to provide high-precision wide-area network communication simulation for the simulation; proposing a dynamic scheduling protocol triggered by events among multiple computing power centers, which achieves causal consistency of cross-domain operations based on logical clocks. The proposal of this architecture makes up for the lack of simulation tools in the context of wide-area multiple computing power centers. It provides efficient and reliable simulation support for relevant researchers in the field of wide-area intelligent computing.
Keywords: multi-computing centre collaboration; wide-area environments; computing network convergence; simulation architecture