how relate population to tail?

Hi, I'd really appreciate if someone could help with the following question that's been bugging me for months...

If I'm working with a zero-mean Gaussian discrete random variable with a standard deviation (sigma) of 1, and I know the population, how far into the tail of this distribution must I go to capture 99.7% (e.g. 3-sigma) of the population?

I know higher populations reach deeper into the tail, I just don't know how to mathematically relate the two.

BR, -GGK

re-state the question: population vs peak-peak value

Sorry, the above question is incorrect -- let me re-state the question:

I'm working with a zero-mean Gaussian discrete random variable with a standard deviation (sigma) of 1.

If I know the population size, what is the peak-to-peak value of the distribution with 3-sigma (99.7%) confidence?

BR, -GGK