STM32F7 maximum speed
Recently, I began to deal with the STM32F7 stone, as the STM32F4 lacked speed. By the result, I realized that if you simply run the same program (calculating digital filters) on F4 and F7 at the maximum core frequency, then the performance does not increase much. But if you enable the first level cache on F7, the calculations are much faster. The question is how this cache is used? Its only 4Kb for RAM and 4Kb for flash. How does the compiler decide what to load there? I also read on F7 that TCM memory can be used to increase performance, but I can’t figure out how to work with it not in Keila, not in stmIDE. Can anyone explain how to work with TCM memory?
