ReRAM: A Strong Scalable Non-volatile Memory
Dirk Wouters
Aachen University / Germany
RRAM memories did catch a lot of research interest because of strong scaling potential, as storage of information is based on atomic arrangement and not on electron charge. Indeed, over the last couple of years, RRAM memories have shown a strong progress towards a promising scaled memory technology, achieving even
While initially people were fascinated that a large group of metal-oxide materials all showed “similar” resistive switching capability, today research focuses on the details of the differences in switching of different material systems. Indeed, where scaling and integration developments focused so far on CMOS-friendly and possibly already-available materials, it is realized that they may not be the best choice. Further physical understanding of the switching process and link to critical material properties are believed to enable RRAM material optimization as a key solution path for RRAM.
Possible applications targets for RRAM are “high-performance” FLASH alternatives (as Storage Class Memories) as well as embedded NVM and memory-in-logic, while further developments of selectorless cell concepts (e.g. using self-selecting devices or complementary switching devices) may enable dense 3 dimensional memory arras that can even compete with current 3D Flash concepts.