Hi,
I am using OMAP L138.
In our application, we have an algorithm that needs to call pow() function. There are ~ 1024 points we need to convert from dBm to watt, it takes ~ 40 ms in total.
double power_watt[i] = pow(10,(power_dBm[i]-30.0)/10.0);
My understading is that ARM has no FPU, so the pow() function (from standard library) is implemented in software.
Any suggestion to make pow() function faster ?
There is a coprocessor CP15 in ARM, not sure can that help ?
Any suggestion will be great ?
rgds,
kc Wong