A Library for Real Time Interactive High Performance Computing of 2D and 3D Physics Problems through GPU*

Many interesting and complex physics problems such as those that use large scale reaction-diffusion equations to study animal skin patterns, brain activity and cardiac arrhtyhmias require the use of supercomputers. Likewise is the case for systems displaying turbulent fluid dynamics, surface growth, and systems using heat and wave equations. We have developed a library using WebGL, which allows to code and run physics problems, that generally require supercomputers to run, in the GPU of a local PC and even a cellphone, which can run up to 7billion differential equations per second.
We have used this library not only for research but also with undergraduates in Computational Physics Courses at Georgia Tech and in workshops at RIT. In this talk we present the library and show interactive simulations of interesting problems in real time. 2D and 3D examples including Turing patterns, turbulent fluid flow, coral growth and fibrillation in the heart among others.