The app we started developing for Caltech last year is very close to be released in Google Play and the App Store. The app uses a mathematical technique named  Reduced Order Modeling (ROM) to build a surrogate one, as introduced in this paper, to substitute months of supercomputer simulations of binary black hole collisions with real-time, online evaluations which produce indistinguishable results.  This technique reduces computational time by 10 or more orders of magnitude and can be used in broad range of very complex problems that would require months or even years of computation in a cluster.  It was developed in collaboration with Jonathan Blackman (Caltech), Chad Galley (Caltech), Scott Field (UMD), Rory Smith (Caltech), Bela Szilagyi (Caltech), Manuel Tiglio (UMD & Caltech), with the LIGO experiment (http://www.ligo.caltech.edu) as a scientific driver motivation. However, the ROM techniques and app are generic and versatile for any problem which requires fast, real-time predictive responses.

Check out the video: http://www.youtube.com/watch?v=4L03Nv86zxs