The formula is also equivalent to :
$$ \sum_{m=1}^N \left \lfloor \frac{N}{m} \right \rfloor $$
An interpretation would be to count the discrete rectangles with total area inferior to N. But aside from that, I don't really see how to express this...
Of course, it should not be hard to compute, but I was curious to see if there is a formula for this.
On the same topic, is the following equality true ? Basically the same formula with bounding both $m$ and $n$ (and not just their product). All variables are integers.
$$ \text{card}\left( \Big\{(m,k)\, s.t.\, m\cdot k \leq N, m\in[1,M] , k\in[1,K] \Big\}\right) = \sum_{m=1}^{\min(M,N)}\min\left(K,\left \lfloor \frac{N}{m} \right \rfloor\right)$$
This is called Dirichlet's divisor problem and is a famous problem in number theory. From Wikipedia:
The formula with quasi-polynomials is not correct. In fact, nobody knows how to derive an explicit formula. Asymptotics have been studied, though.