This paper investigates the problem of Bias Temperature Instability (BTI) and Hot Carrier Injection (HCI) on processors. We propose a performance- and reliability-aware methodology that evaluates the effects of these degradation mechanisms at design time. The performed analysis estimates the effects produced by the execution of applications, representing typical or worst case scenarios, or single instructions. As shown by the experimental results, our framework allows to estimate the performance degradation and to identify the areas of memory most subject to faults, with the objective of optimizing the system design and defining on-line strategies.