I wrote a simple code to compare the speed of Python vs. C++. Here is the Python code:

def count_permutations(N:int)->int:
    count=0
    for p in itertools.permutations(range(1,N+1)):
        count += 1
    return count

And here is the C++ code:

int count_permutations(int N) {
    vector<int> v(N);
    iota(v.begin(), v.end(), 1); // fill the vector with 1,...,N
    int count=0;
    do {
        ++count;
    } while ( next_permutation(v.begin(),v.end()) );
    return count;
}

When I run with Python 3.8.5 on Ubuntu, I get:

Permutations of 1..11:
  39916800 permutations calculated in 3.31172251701355 seconds
Permutations of 1..12:
  479001600 permutations calculated in 40.63520336151123 seconds

When I run with C++ (clang++-9) on the same Ubuntu, I get:

Permutations of 1..11:
  39916800 permutations calculated in 5 seconds
Permutations of 1..12:
  479001600 permutations calculated in 57 seconds

I find this very strange, since in the benchmarks website, C++ is always faster than Python.

Is it possible that recent versions of Python have become so fast that they are now faster than their C++ parallel?