Speeding up Static Probabilistic Timing Analysis

Probabilistic Timing Analysis (PTA) has emerged recently to derive trustworthy and tight WCET estimates. Computational costs due to the use of the mathematical operator called convolution used by SPTA – the static variant of PTA – and also deployed in many domains including signal and image processing, jeopardize the scalability of SPTA to real-size programs. We evaluate, qualitatively and quantitatively, optimizations to reduce convlution’s computational costs when it is applied to SPTA. We showthat SPTA specific optimizations provide the largest execution time reductions, at the cost of a small loss of precision.