@ID_AA_Carmack perhaps, but single thread mobile perf is nowhere *near* as optimized already as, say, Sandy Bridge -> Ivy -> Haswell