The latter is right. 2^n can never end on a zero; The last digits of something that is doubling behaves like this:
Graph[Table[n -> Mod[2 n, 10], {n, 0, 9}], VertexLabels -> "Name"]
as you can see, the last digit will go through a cycle: 2,4,8,6,2,4,8,6......... which you can check for the first 2^n:
2,4,8,16,32,64,128,256,512,1024,2048,...
indeed that matches. This does not mean the wolframalpha is correct, but is certainly proofs that excels and Open Office Calc are incorrect. (mainly because they use machine precision numbers and they can not contain all the digits (only 53 binary digits for a double-precision-number if I recall correctly)).
