I don't know if it's still the case, but when I was an undergrad a professor pointed out that Excel's Gaussian tail probabilities disagreed with everything else

I don't know if it's still the case, but when I was an undergrad a professor pointed out that Excel's Gaussian tail probabilities disagreed with everything else (Mathematica, Matlab, R, etc.) for large values.