Cluster-based delta-QMC technique for fast yield analysis

Nguyen Cao Qui, Si Rong He, Chien-Nan Liu*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

Monte Carlo (MC) analysis is often considered a golden reference for yield analysis because of its high accuracy. However, repeating the simulation hundreds of times is often too expensive for large circuit designs. The most widely used approach to reduce MC complexity is using efficient sampling methods to reduce the number of simulations. Aside from those sampling techniques, this paper proposes a novel approach to further improve MC simulation speed with almost the same accuracy. By using an improved delta circuit model, simulation speed can be improved automatically due to the dynamic step control in transient analysis. In order to further improve the efficiency while combining the delta circuit model and the sampling technique, a cluster-based delta-QMC technique is proposed in this paper to reduce the delta change in each sample. Experimental results indicate that the proposed approach can increase speed by two orders of magnitude with almost the same accuracy, which significantly improves the efficiency of yield analysis.

Original languageEnglish
Pages (from-to)64-73
Number of pages10
JournalIntegration, the VLSI Journal
Volume58
DOIs
StatePublished - 1 Jun 2017

Keywords

  • Monte Carlo analysis
  • QMC
  • Yield analysis

Fingerprint Dive into the research topics of 'Cluster-based delta-QMC technique for fast yield analysis'. Together they form a unique fingerprint.

Cite this