Abstract
Reliability evaluations are conducted to determine if non-standard material meets a company's reliability requirements and whether or not it may be sold to customers for revenue as normal material. Material is considered to be non-standard if it was misprocessed, fails to conform to specifications or any criterion deemed to critically affect the performance or reliability of the device, or demonstrates characteristics that result in a substantial concern about the reliability of the product. A methodology for establishing sampling plans for reliability experiments designed to evaluate non-standard material is discussed in this paper. Two case studies from Intel Corporation are shown.