The goal of this paper is to describe some of the design choices made in a BladeCenter system design. This paper will focus on the connector modeling that was used to make design trade-offs, unique to a 1OGbps serial system. Connectors are a significant source of crosstalk noise in a system, which could increase jitter in the resulting eye pattern. Since crosstalk jitter is uncorrelated from the link data pattern, it is hard to remove through equalization techniques. This paper provides a process to extract accurate coupled models from connectors. It also describes how modeling was used to predict crosstalk noise and shows how a small percentage of connector crosstalk can have a significant impact on the over all signal distortion, if the connector pin assignment is done randomly IntroductionDesign choices in a Blade Center backplane system are based on simulation models. Accurate simulation models are critical to making good design trade-offs, especially in 1OGbps serial systems. Typically, models for several components provided by external suppliers have limited accuracy. As serial link speeds have increased, minimizing crosstalk has become a major Signal Integrity challenge. In a 1OGbps system connector crosstalk can be a significant jitter contributor if the connector pin assignment is done randomly. However, the effect on the overall system can be reduced by following certain design guidelines.This paper provides a process to extract accurate connector crosstalk models. The Molex SEARAYTM connector was evaluated to determine its suitability in the high speed backplane system. As shown in Figure 1 below, the BladeCenter interconnects are card to card high speed channels, which enable system compatibility between shared devices. The backplane in a blade system can be defined as a flexible interface to attach shared devices such as processor blades, switches, management modules, etc.Two different techniques were used to measure the near end and far end crosstalk contribution of the connector. The correlation of crosstalk measurement techniques was done using a VNA in the Frequency domain and a TDR in the Time domain. This procedure validated the connector crosstalk measurements, which with a high confidence to see the impact of connector cross talk.In addition to characterizing the crosstalk contribution of the connector by itself, a channel simulation using the StatEye tool was performed using the validated coupled models that were created by exporting measured frequency domain data into the touchstone format.