Sidda Reddy Kurakula¹, Lokesh Kulkarni¹, Madhu Dasari¹, Helen Armer², ¹Applied Materials India (P) Ltd, India and ²Applied Materials, Inc., USA

Integrated circuits (a.k.a chips or IC’s) are some of the most complex devices manufactured. Making chips is a complex process requiring hundreds of precisely controlled steps such as film deposition, etching and patterning of various materials until the final device structure is realized. Also, each chip goes through a huge number of complicated tests and inspection steps to ensure quality. In IC manufacturing, yield is defined as the percentage of chips in a finished wafer that pass all tests and function properly. Yield improvement translates directly into increased revenues. A humongous amount of data (Terabytes per day) is logged from the equipment in the fab. This paper describes some applications of advanced data mining techniques used by chip makers and equipment suppliers in order to improve yield, match equipment, increase equipment output and also to predict the change in equipment performance before and after maintenance activities

Integrated Circuit (IC), Yield, Equipment, Wafer, Models