By the time we showed SDS 2 to other customers, it had already been deployed at GlobalFoundries FAB 8 in Malta. We held a large workshop with over 30 people. The room was very crowded with many interested and skeptical people listening to my presentation about SDS 2.
After the workshop, there were many questions like: What if we are looking for signatures only in this area on the wafer? This answer was easy because SDS 2 has a zoning tool to mask areas on the wafer to look for signatures.
But then suddenly another question followed: What about, we are looking for signatures in this area up there and if there is nothing found we would like to lookup this area and if ... You can image how this line of questioning turns out. In this workshop I got so much input from the engineers and their different use cases that it was hard to satisfy all of their needs.
Additionally, there were some workflow improvements needed, because we had a new algorithm to detect concentric rings which was no longer relying on a pre-cluster step.
We moved forward, I spoke with a lot of people and interested customers for our SDS 2 product. After some time I realized SDS 2 was not enough to satisfy the broad range of needs for a modern semiconductor manufacturing site. I began to think about a deeper meaning in a defect solution.