Crosstalk analysis has become a significant part of the design cycle of high performance processors in nanometer technologies. In this paper we demonstrate that current crosstalk analysis techniques that ignore the degrading effect of multiple crosstalk events on receiver noise rejection curve filter significant number of true violations. We also demonstrate that techniques that take into account the multiple crosstalk events with traditional receiver modeling result in large number of false violations. We propose improved crosstalk analysis techniques that are multiple noise event aware (MNEA) with minimal changes to existing crosstalk analysis. We also propose enhancements to existing receiver models so they can be used with the MNEA analysis resulting in reduction of number of false violations by 68%-98% while guaranteeing identification of all true violations. © Springer-Verlag Berlin Heidelberg 2006.
CITATION STYLE
Nanua, M., & Blaauw, D. (2006). Receiver modeling for static functional crosstalk analysis. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4148 LNCS, pp. 329–339). Springer Verlag. https://doi.org/10.1007/11847083_32
Mendeley helps you to discover research relevant for your work.