Hierarchical power delivery network analysis via bipartite markov chains

This paper proposes a Markov-chain-based hierarchical method to efficiently analyze the power delivery network. After the network is partitioned into several sub-networks, each sub-network is transformed into a local Markov chain. Then, the connectivity between all the sub-networks is modeled as a global Markov chain by using a novel technique called parallel-street-walk. Finally, these local and global Markov chains are integrated in order to build a hierarchical bipartite Markov chain engine to analyze the power delivery network. The experimental results not only demonstrate the accuracy of our proposed method when compared with a very accurate time-domain solver, INDUCTWISE [2], but also show a significant runtime improvement- over 200 times faster than the INDUCTWISE and over 10 times faster than the IEKS method [3] - and low memory consumption.