Performance modeling of resonant tunneling-based random-access memories

Abstract

Resonant tunneling-based random-access memories (TRAMs) have recently garnered a great amount of interest among memory designers due to their intrinsic merits such as reduced power consumption by elimination of refreshing operation, faster read and write cycles, and improved reliability in comparison to conventional silicon dynamic random access memories (DRAMs). In order to understand the precise principle of operation of TRAM memories, an in-depth circuit analysis has been attempted in this paper and analytical models for memory cycle time, soft error rate, and power consumption have been derived. The analytical results are then validated by simulation experiments performed with HSPICE. These results are then compared with conventional DRAMs to establish the claim of superiority of TRAM performance to DRAM performance.