Chip parameter yield prediction method considering multiple property constraints at same time

Abstract

The invention discloses a chip parameter yield prediction method considering multiple property constraints at the same time, and belongs to the technical field of multi-element parameter yield prediction under the condition that chip design parameter perturbation has random correlation. Current chip parameter yield research is mainly restricted to predicting single property index yield or carryingout balanced optimization on a plurality of single-property index yields. Aiming at the problem that above method cannot consider correlation between multiple property indexes so that parameter yieldloss is easy to cause, the invention discloses an accurate chip multi-element yield prediction method considering multiple property constraint index influences at the same time. The method specifically comprises the following steps of: firstly, considering correlation between design parameter perturbations and constructing a chip property statistic model; secondly, solving an edge distribution probability of each chip property by utilizing a saddle point line sampling method; and finally, selecting an optimum Copula function by utilizing a Copula theory so as to accurately predict a chip parameter yield considering multiple property constraints at the same time.

Description

technical field [0001] The invention relates to a method for accurately predicting the yield of chip parameters aimed at randomness of design parameter disturbance, and belongs to the technical field of computer-aided design. Background technique [0002] At the current level of chip manufacturing technology, the deviation of chip performance from the design index caused by the disturbance of design parameters will lead to a significant decline in the yield of chip parameters. In particular, since design parameter disturbances usually act on different chip performance indexes at the same time, there will be a strong correlation between chip performance indexes. Therefore, it is very important to accurately predict the yield of chip parameters by considering the impact of multiple chip performance indicators at the same time. [0003] So far, most of the known chip yield prediction methods are limited to the yield prediction for a single performance constraint index, while i...

AI Technical Summary

Problems solved by technology

This method considers the accurate prediction of the yield of chip parameters under the simultaneous action of multiple performance constraints, so as to solve the problem that in the process of computer-aided design, chip designers can consider the strong correlation between various chip performance indicators, and the disturbance of design parameters Accurately predict the yield of chip parameters affected by randomness

Images