A Microarchitectural Analysis of Soft Error Propagation in a Production-Level Embedded Microprocessor

Current trends in device scaling continue to cause an increasing risk of transient faults in microprocessors due to high energy strikes from radiated particles. In this work, we present a thorough microarchitectural analysis of the effects of soft errors on a production-level Verilog implementation of an ARM926EJ-S core. We examine the propagation of faults occurring in both sequential state elements and combinatorial logic and note a number of critical distinctions in the error propagation behavior of soft errors occurring at logic gates versus state elements. Further, we exemplify the ways in which the emerging trend of faults in combinatorial logic will affect the scope of the soft error problem, especially in the embedded design space. Also, since this work was conducted on a production-level core, we highlight some of the nuances of soft error effects that arise and are specific to production-level designs.