Reference: Mohammed, J. & Losleben, P. Applications of Artificial Intelligence to Semiconductor Modeling. Knowledge Systems Laboratory, October, 1993.
Abstract: Most of the research performed over the past twenty years in semiconductor process modeling has relied on fairly conventional methods of scientific programming. While these are adequate for many simulation and analysis applications, the more complex problems of synthesis and diagnosis suggest the need for alternative approaches to modeling. This paper examines the breadth of potential applications of artificial intelligence (AI) to synthesis and diagnosis. Successful application of AI to this field is sparce, but enough specific examples exist to suggest high potential for future use of the approaches that are described. The paper identifies tasks in semiconductor modeling which are candidates for application of AI methodology, catalogs appropriate methodology and surveys applications at the current state-of-the-art.
Notes: submitted to IEEE Trans. Semicond. Manuf. Stanford Integrated Circuits Lab., Manuf. Sci Tech. Note Series No. 5.
Full paper available as ps.