How to speed the code in msp430F47177 microcontroller

Gokul Tile said:

there is large number of modulo & devide operations of float data types.So execute such float modulo & divide operation microprocessor requires more execution time

So don't use floating point, then!!

• With suitable scaling, you may be able to work entirely in integer maths; eg, instead of 1.23V, use 1230mV;
   
• Have you considered fixed-point?

 

Andy Neil said:

With suitable scaling, you may be able to work entirely in integer maths; eg, instead of 1.23V, use 1230mV;

Indeed, that's by magnitudes faster. Also, often a formula can be optimized so it doesn't use division, or uses multiplication and a shift.
Using float (or double) is in most cases totally unnecessary. What good is a result with 10 digits after decimal point if the error in the parameters in before the decimal point (and often not on the lowest digit).

The only problem with scalign is that you need to take care of the possible value size in intermediate results. The formula must be reordered so that you never exceed the allowed value range. So firs tmultiply and hten divide, when dividing first would bring your value so low that the integer truncation removes valuable fractions. Or divide first, if the multiplication would exceed the int/long int range. It needs a bit 'brainwork', but but better spending some time at development than continuously wasting precious realtime cycles on unnecessary float operations.