Compiler: when we have Float then why we _iq?!

Dave Hunson

Expert 1740 points

Tool/software: TI C/C++ Compiler

hi every one !

why we have to use _iq when we have float?

what is _iq ?how _iq types works?

thanks for attention

regards

Dave.

over 6 years ago

0 Keith Barkley over 6 years ago

Guru 42215 points

If you don't have an FPU, floats take a lot of CPU time. Even the multiply/divide only FPU like in the MSP432 might make some operations take a long time. Using integer math is much faster in that case, and can be tuned for word size and memory access.

iq is fixed based math. You just set a decimal place. For example, your 16 bit integer number will get you from about -32768 to + 32768. If you are working with US money, you might want two decimal places so your data ranges from -327.68 to +327.68. You do all your work with the integers and just supply the decimal point when you display it, This is just a toy idea, and there are some other subtleties, but you get the idea.

0 Dave Hunson over 6 years ago in reply to Keith Barkley

Expert 1740 points

thanks dear I got some thing about that ,so what is the principle of it?
let say float MyVar= 12.4500; then how to convert it to _iq24?and vice versa?
I don't want to use function or macros I want to know the principle of _iqx...

thanks for attention.
Dave.

0 Keith Barkley over 6 years ago in reply to Dave Hunson

Guru 42215 points

processors.wiki.ti.com/.../IQMath_fixed_vs_floating.pdf

I believe you would multiply by 2^24 in floating point and then convert it to an integer.

0 Dave Hunson over 6 years ago in reply to Keith Barkley

Expert 1740 points

Thanks dear :)
It helps a lot !
Dave.

Code Composer Studio™︎

Code Composer Studio forum

Compiler: when we have Float then why we _iq?!