Work out wafer yield (murphy model) instantly with clear inputs, formula shown and shareable results.
Murphy's model assumes defect density varies across the wafer, giving Y = ((1 - e^-AD)/AD) squared. It sits between the pessimistic Poisson model, which assumes uniform random defects, and the negative binomial model that explicitly parameterises clustering. All three show yield collapsing as die area grows.
Murphy yield
Y = ((1 - e^(-A D0)) / (A D0))^2, with A in cm2 and D0 in defects per cm2
Because defects cluster. Clustering means some regions carry many defects while others are clean, so more die escape entirely than uniform statistics predict.
Yield falls roughly exponentially with area. Doubling die area can easily halve yield, doubling cost per good die on top of the halved die count.