simple calculation using logs

Suppose we are comparing implementations of insertion sort and merge sort on the same machine. For inputs of size $n\in\mathbb{N}$, insertion sort runs in $8n^2$ steps, while merge sort runs in $64n\log(n)$ steps. For which values of $n$ does insertion sort beat merge sort?

-
Note: n belongs to the natural numbers – Tom Mar 1 '12 at 17:42
It's equivalent to solving $n<8\log(n)$. Assuming the logarithm is 2-based, if $2\leq n\leq 32$, the inequality holds. The exact range can be approximated numerically. – Hawii Mar 1 '12 at 17:50

Solving $8n^2 \lt 64n\log(n)$ (equivalently, $n\lt 8\log(n)$) exactly requires use of Lambert's W function. Solving it approximately is pretty easy.
At $n=1$, $8n^2\gt 64n\log(n)$; at $n=2$, $8n^2 \lt 64n\log(n)$; if $\log$ is base $2$, then they cross again at about $n=45$.