Case study of yield learning through in-house flow of volume diagnosis

To find out the root causes of yield loss is always expensive and time-consuming. In this paper, we have developed an in-house flow of volume diagnosis. With the power of performing statistical analysis on the large volume of diagnosis results, we demonstrate that to do physical failure analysis could be in a more economical way. By utilizing the in-house flow of volume diagnosis, we could find out the defects which cause the yield loss primarily with only a few samples of failing dies in a single iteration. The industrial cases demonstrate the efficiency proved by the silicon data.