Ah my mistake then. I didn't realize the closed form solution takes that long to evaluate.

Couldn't there be a faster way to evaluate it using an algebraic rather than a computational way? I just find the bound you give too high but I may be completely wrong.