Analytical approaches for worst and statistical eye estimation in HBM channel

Abstract

Analytical approaches are proposed for the estimations of the worst case and statistical eye diagrams in high bandwidth memory (HBM) channels. The target HBM channels consist of the transceiver (Tx) drivers, symmetric coupled channel, and capacitance loads. The proposed estimation methods include the Tx power noise and channel crosstalk in their estimation procedures. Based on the proposed analytical approaches, we can obtain the fast estimation time for the eye diagrams and the physical insight into the distortion in the eye diagrams caused by the Tx power noise and channel crosstalk. The proposed analytical approaches are validated by comparisons with HSPICE simulations.