Space Optimization

Notice that to compute fib(i), you only need fib(i-1) and fib(i-2). You don't need the entire array. So instead of storing all $n+1$ values, you can keep just two variables: prev1 (for fib(i-1)) and prev2 (for fib(i-2)). At each step, you compute the new value, then shift: prev2 = prev1, prev1 = current.

This reduces space from $O(n)$ to $O(1)$. This "rolling window" technique is common in DP when each value depends only on a few recent values and saves you tons of memory.