Logic BIST: State-of-the-Art and Open Problems

Abstract

Many believe that in-field hardware faults are too rare in practice to justify the need for Logic Built-In Self-Test (LBIST) in a design. Until now, LBIST was primarily used in safety-critical applications. However, this may change soon. First, even if costly methods like burn-in are applied, it is no longer possible to get rid of all latent defects in devices at leading-edge technology. Second, demands for high reliability spread to consumer electronics as smartphones replace our wallets and IDs. However, today many ASIC vendors are reluctant to use LBIST. In this paper, we describe the needs for successful deployment of LBIST in the industrial practice and discuss how these needs can be addressed. Our work is hoped to attract a wider attention to this important research topic.