Optimizing Zingg Performance for Processing 90 Million Records: Solutions and Data Splitting Options

Hey, We have 90M records as input to Zingg, its taking too much time (more than 11hrs), Please provide better solution approach to this ? can we split data ? or anything!!