well, as a practical matter that limit isn't used computationally, since it doesn't converge as fast as other expressions for e.
some approximations for a few values of N:
N = 1:
(1 + 1/1)^1 = 2
N = 2:
(1 + 1/2)^2 = 9/4 = 2.25
N = 10:
(1 + 1/10)^10 = (11/10)^10 = (25,937,424,601)/(10,000,000,000) = 2.5937424601
N = 100:
(1 + 1/100)^100 = (101/100)^100 = 2.7048138294215260932671947108075 (or so)
as you can see, the exponents get pretty big, making calculation difficult.
a better way to find a value for e, is using the taylor series for e^x centered at 0:
e^x = 1 + x + x^2/2! + x^3/3! + ....+x^n/n! +......
using x = 1, this becomes:
e = 1 + 1 + 1/2 + 1/6 + 1/24 +......+ 1/n! +.......
so for n = 1, we get the approximation 2
for n = 2, we get 1 + 1 + 1/2 = 2.5
for n = 3, we get 1 + 1 + 1/2 + 1/6 = 8/3 = 2.66666.....
for n = 4, we get 1 + 1 + 1/2 + 1/6 = 65/24 = 2.708333......
(it took N = 100, to get this close with the other limit).
for n = 5, we get 326/120 = 2.7166666......
for n = 6, we get 1957/720 = 2.71805555......
which is accurate to 3 decimal places.
you can carry out the terms as much as you have patience for,
but factorials get big very fast (so their reciprocals get small very fast).