There are 2 basic types of data deduplication: In-line and post-process.

In-line deduplication is a process that categorizes duplicate data in real-time. With post-process, the task of deduplicating is performed at pre-determined intervals.

This one-page transcript of a presentation given by deduplication expert Howard Marks gives an overview of the pros and cons of different dedupe processes, and delves into the type that Microsoft's Hyper-V 3.0 employs and why.