Deep learning is clearly a critical success factor for any semiconductor business, whether a fab, mask shop, equipment or service supplier, or a fabless customer. In the deep learning era, data is everything so protecting the data you have is critical. And what about the data you don’t have but need? In electronics manufacturing, the relatively few defect occurrences in mask and wafer fabs are insufficient for training a deep learning neural network. The solution is to create a digital twin, a simulated version of the real thing whether it’s hardware or software. From design-to-silicon, everything will need a digital twin. Let’s start working on it now.