Date of this Version
The reject rate of LSI chips due to incomplete fault coverage of the tests is the fraction of faulty chips, among the chips that pass the tests. This reject rate, which is a measure of the tested chip quality, contributes to the field returns. It is, however, difficult to determine the tested chip quality from the field return data which may also include rejects due to handling damages, infant mortality, etc. Also, a large number of chips must be in use in the field before an adequate amount of field return data can be obtained. This paper gives a method of forecasting the reject rate from the test data alone before any field trials are made.