Mike Plisinski, VP/GM, Data Analysis and Review Business Unit, Rudolph Technologies, Inc., Bloomington, MN USA

With our industry in a cyclical downturn exacerbated by the worldwide financial crisis, we are challenged to do everything possible to maintain profitability. Unfortunately, times such as these can also increase aversion to perceived risk, resulting in missed opportunities for incremental but significant gains. It is tempting to adopt an “If it ain’t broke, don’t fix it” attitude. We would do well not to neglect the lessons of the automotive industry in which competitors who embraced a philosophy of “constant improvement” demonstrated once-and-for-all the superiority of that approach over one that aspired only to be “good enough.”

To some extent, the aversion to risk also reflects the extraordinary complexity of the semiconductor manufacturing process where unintended consequences seem to be the rule rather than the exception. The exhaustive analysis required to identify opportunities for marginal gains and confidently prescribe changes in procedures that are already delivering targeted yields is difficult to justify. The situation is made worse by the compartmentalization of performance analysis within the fabrication process where groups are evaluated on the performance of their module rather than the overall performance of the entire manufacturing process.

The advent of powerful and accurate software systems capable of automated modeling and analysis of the entire complex manufacturing process helps engineers to maximize yield and lowers the risk in doing so. Engineers are able to identify root causes of yield loss, consider suggestions for corrective action and obtain quantitative predictions of resulting gains. The value these programs contribute can easily make the difference between success and failure in these challenging times.