ISBN-13: 9781402072550 / Angielski / Twarda / 2002 / 250 str.
ISBN-13: 9781402072550 / Angielski / Twarda / 2002 / 250 str.
Design and test are considered jointly in this book since knowledge of one without the other is insufficient for the task of having high quality memories. Knowledge of memory design is required to understand test. An understanding of test is required to have effective built-in self-test implementations. A poor job can be done on any of these pieces resulting in a memory that passes test but which is not actually good. The relentless press of Moore's law drives more and more bits onto a single chip. The large number of bits means that methods that were "gotten away with" in the past will no longer be sufficient. Because the number of bits is so large, fine nuances of fails that were rarely seen previously now will happen regularly on most chips. These subtle fails must be caught or else quality will suffer severely. Are memory applications more critical than they have been in the past? Yes, but even more critical is the number of designs and the sheer number of bits on each design. It is assured that catastrophes, which were avoided in the past because memories were small, will easily occur if the design and test engineers do not do their jobs very carefully.
High Performance Memory Testing: Design Principles, Fault Modeling and Self Test is based on the author's 20 years of experience in memory design, memory reliability development and memory self test. High Performance Memory Testing: Design Principles, Fault Modeling and Self Test is written for the professional and the researcher to help them understand the memories that are being tested.