Answers are grounded in this paper and cite their sources.
In order to make sure we have the infant-mortality problem in perspective, let us take a look at Fig. 9. This shows the typical infant-mortality failure rate in digital bipolar integrated circuits compared to those resulting from the failure mechanisms typical of the main population. The range of the long-life main-population rates is suggested to be within realistic expectation, based on accelerated-stress tests on a variety of integrated-circuit types in recent years. It represents product which could pass a typical commercial life test of 1000 hours at 125°C. What is demonstrated is that, even though the long-life reliability of an integrated circuit could be quite satisfactory, infant mortality can provide failure rates which could be undesirable for many present-day equipments, for times which are significant with respect to equipment life. For example, the calculations from MIL-HBK-217 provide failure rates of SSI-MSI digital integrated circuits as about 10 to 100 FIT's, yet we see in Figure 6 that this failure rate is achieved only after from one month to one year of equipment operation.
No takes yet. Share an insight, caveat, or question.
D. S. Peck (1978) studied this question.
Synapse has enriched 9 related papers. Consider them for comparative context: