Hello everyone,
I have an i.MX6 SabreSDB (27392 RevB5, 700-27392 Rev X11) board running @ 996 MHz. I need a better performance on some algorithms and so being able to set the CPU clock @ 1.2 GHz. I am using the Dora Branch of the Yocto project.
I know there are already some posts on the Community about particular point (cpu clock, or https://community.freescale.com/message/355102#355102) but I do not really understand them because I am not from Hardware. I read that it could be possible with the "Fuse" but I am not really confortable with that. Please, can someone give me a methodology or some understandings about that ?
Moreover what are the pros/cons of such a clock modification ?
Regards,
Mathieu
Mathieu
This discussion is closed since no activity. If you still need help, please feel free to reply with an update to this discussion, or create another discussion.
Thanks,
Yixing
Mathieu
Had your issue got resolved? If yes, we are going to close the discussion in 3 days. If you still need help, please feel free to reply with an update to this discussion.
Thanks,
Yixing
BTW, Which are the areas you want to improve performance?
Hello,
I am working on C/C++ implementation of image processing algorithms. The implementation has already been tune to fit the ARM Cortex A9 but I still need the code to run 15% faster.
Regarding the performance @ 792MHz, boosting the clock to 1.2 GHz should provide me the overall performance I need for my application.
Regards,
Mathieu
Is something can be done using the IPU? even if you increase the CPU's clock, the IPU will be much faster for this processing area, so make use of it if possible.
No, those image Processing algorithms are really complex. The IPU cannot be helpful here.
I see. In case you are doing matrix/vector calculation, be sure to enable neon when compiling.