Nvidia’s GPU failures: A case for prognostics and health management

Prognostics and health management (PHM) is an enabling discipline consisting of technologies and methods to assess the reliability of a system in its actual life cycle conditions to determine the advent of failure and mitigate system risks. This paper presents a case study of a failure that had a significant economic impact on the computer industry and its customers, and then discusses how PHM implementation could have resulted in a dramatic difference in the outcome.