3 ms·

I'm confused why you wouldnt just do some random sampling to get some statistical bounds. At least then you'll know if you are close.