Would be interesting to time the performance of each. I think sprintf has a pretty hefty toll IIRC
There is also a general risk with the approach you take because it would be possible that v2 x 100 does not fit into a signed int (not in that case as we know v2 is less than 20.46)