The in-memory computing technique and its application in Big Data processing are described first. To address the hard disk and network data I/O bottleneck issues in power system analysis

an in-memory computing technology based power system analysis architecture is proposed. Typical in-memory computing application scenarios in the fast DSA implementation are discussed. A power network modelis used as an example

which is available from the online DSA system currently demonstrated in the national power dispatching center in China. The corresponding performance test results of in-memory computing in large power network analysis are presented.