Big is a relative term. Data is pretty general.
Ryan Shuttleworth said yesterday at the Turing Festival that “big data forces you to innovate about how you store it and process it”. On the presentation display there was a number line going from 100GB to 1EB (or something utterly huge).
The exact behaviour will vary by application and the data model, but basically all algorithms and storage methods will work well enough on tiny data sets. It’s when the data starts to grow that you’ll notice problems at scale.
At Microsoft’s SQL Server 2012 launch in Edinburgh, the presenter talked about big data as having the four V’s: volume, velocity, variety, and variability. But that’s just replacing one relative term with four others.