When you mentions billion row datasets MapReduce and Apache Hadoop comes to mind, but that requires that you are capable to do some computer programming.

There may also be a lot of existing solution to present/summaries/graph you data, depending on what it contains and witch program created it. Can you give us some more insight into what kind of data you have?