Asymptotically faster means that eventually it grows larger, but doesn't strictly mean that it's always going to be faster at the beginning. i.e. n^2 grows asymptotically faster than 10*n as n goes to infinity. For n < 10, n^2 < 10*n, but for n > 10, n^2 > 10*n. This is what big-O is about, eventually O(n^2) will grow faster than O(n).
NOTE: if an algorithm X is asymptotically Faster than other algorithm Y for same set of input, it means X will take more time than Y so, Y is efficient than X.