Statistical learning (SL) is routinely implemented in enterprise systems for search, data mining, pattern recognition, etc., by large corporations such as Google and most recently Apple, with its introduction of its voice-controlled assistant Siri. SL is typically employed when the problem of concern cannot be easily modeled. More recently, SL is being used to address many of the complexities inherent to systems that are engineered by the semiconductor industry. SL for efficient, holistic optimization of integrated-system operation however requires new learning algorithms since conventional users of SL assume external, virtually limitless compute and storage resources. Integrated systems, especially mobile systems, have stringent constraints on power that necessitates a fundamental re-thinking of how to implement learning especially since the system itself both performs the learning and uses the resulting knowledge. Therefore, it is conceivable that some learning tasks will require custom hardware, for others software or even cloud-based solutions may suffice. In this project, we propose to demonstrate the capability of SLIC through the development of a K-nearest neighbors hardware architecture for the mitigation of system aging.