PROJECT TITLE :

Silent Data Corruption and Embedded Processing With NASA's SpaceCube

ABSTRACT :

Dramatic will increase in embedded data processing performance are turning into attainable using platforms such as the NASA SpaceCube. With a versatile architecture and industrial devices, selected computations will be tuned for the highest performance while let alone perfect data reliability. More needs to be known about the character of silent information corruption during this paradigm. When it occurs, how pervasive is it? To what extent can it's mitigated whereas near-optimal performance is maintained? This paper provides new insights into these queries, via a fault emulation-based study of 2 disparate applications running on a hard-core embedded processor. Two very low-value ways of knowledge error detection reduce the worst sort of silent information corruption (SDC) by eighty nine–ninety seven% with a performance overhead of $< 1hbox%$.