Power analysis on chipwhisperer husky

I am trying to collect power traces using CWHUSKY (target SAM4S), I am multiplying the same numbers multiple times, and getting different traces

You will never get the exact same trace twice because of noise. The ChipWhisperer platform aims to minimize noise, but it cannot eliminate it. Did you have a specific question beyond that?

There is shift 500 to 750…different colors for different iterations…same operation and operands

Have a look at your assembly; your compiler may be playing tricks on you. You may also run into differences stemming from wait cycles for memory access.

For the same code I was getting better results on CWNANO…what exactly should I do? I am trying to see relation between hamming weights and power.

I answered this above. Different targets will, naturally, behave differently!