As noted above, it's not CMU's fault it took so long. They went to Marvell early on (2002?) and tried to license it, Marvell refused and said they found a different way around the patent. It was about 2006 and CMU got some good intel that Marvell was using the patent and restarted negotiations. Those fell through after over a year, and the suits started. Discovery was slow as is usual (and disastrous for Marvell -- look on the web for full details), and the trial took a long time to prepare. They've been slowly grinding through the system, and the fact that they've been slowly grinding through the system means that the award keeps going up because of the magic of compound interest and something that had willful infringement.
Compounded treble damages are something nobody every wants to see. It's why most working engineers are told to be very, very, very careful on anything to do with patents. It's actually better if you never do a patent search because willful infringement is so damaging to the company.