Amid the shift towards more complex chips at advanced technology nodes, TSMC keeps exploring methodologies to discover insightful information among the big data to ensure quality and solve significant production challenges.
Due to the large amount and complexity of data, machine learning nowadays plays an important role in data analysis. These analyses include process parameters optimization, defect inspection classification, materials and quality control, predictive equipment maintenance, and reliability improvement. In many cases at TSMC, we face the need to design a high-dimensional system to find relationships between phenomena and features.
In this talk, we will share our Q&R AI development plan to help us determine the key features and identify some potential risks. We will also demonstrate some cases, which show the benefit of applying machine learning to improve our production quality.