c++ - Different results on using sqrt() and pow() functions

Question

I was solving Euler problem no. 10 that asks to find sum of all primes below 2million. I am getting different results on using sqrt and pow functions. Using sqrt gives the correct answer, plus using pow function takes more time. Here is my code.

for(sum=0,i=3;i<=2000000;i+=2)
{
    for(j=3;j<=sqrt(i);j++)
        if(i%j==0)
            break;
    if(j>sqrt(i))
        sum+=i;
}
sum+=2;
std::cout << "
Sum = " << sum;

Answer 1

I can not reproduce your problem on my computer, but your testing on double is a very dangerous one. If i is the square of a prime number you are relying on high precision of the sqrt and pow to have the correct result. Maybe on your system there is a slight rounding difference on one or more of such squares. You'd better test j*j <= i and j*j > i.