Hi all, I did some of the investigation based on comparison of FPU based algorithms on CM4 and CM7 cores. All the code/data were placed into the single cycle memory with full utilization of modified / true Harvard architecture, it means: - on CM4 -
↧