6 ms·

Will this be efficient for sparse matrices ?