Abstract
High-density components and process scaling lead more and more to the occurrence of new class of dynamic faults, especially in Static Random Access Memories (SRAMs), thus requiring more and more sophisticated test algorithms. Among the different types of algorithms proposed for testing SRAMs, March Tests have proven to be the most performing due to their low complexity, their simplicity and regular structure. Several March Tests for dynamic faults have been published, with different fault coverage. In this paper we propose March BDN, an extended version of the March AB. We will prove that it is able to increase the fault coverage in order to target latest dynamic faults. We show that the proposed March BDN has the highest known fault coverage compared to March Tests with the same complexity.