56800EX math library

[Ben]

Hello

is it normal that the 56F83783 using is so slow in 64bit calculation?

sin calculation for float takes 80usec, (32bit library)

sin calculation for double takes 500usec. (64bit library)

if there a way to fasten the process?

ben

[Celeste_Liu]

Hello Ben,

Thanks for your post. The 56F83783 uses 56800EX core which has no FPU. It can only perform fixed-point operations and does not support floating-point. It has to rely on software simulator for floating-point operations, which is why it is slow.

We recommend the 56F80xxx series.

MC56F80xxx Digital Signal Controllers | NXP Semiconductors

It uses the 56800EF core which has FPU and CORDIC and supports floating-point.

Hope it can help you.

Best Regards,

Celeste