Supplementary Materials to Adaptive and Transparent Cache Bypassing for GPUs

This document is the supplementary supporting le to the corresponding SC-15 conference paper titled Adaptive and Transparent Cache Bypassing for GPUs. In this document, we rst show the experiment gures for the four extra GPU platforms that cannot t into the original paper due to page limitation. We then show the simulation results for the hardware approach that attempts to reduce bypass overhead. Finally, we analyze the performance patterns of the applications with respect to dierent bypassing threshold, which may explain why certain applications can benet signicantly from cache bypassing than others.