We are a team of researchers working in High Performance Computing in Indian Institute of Science, Bangalore. We regularly work on application optimizing applications for GPU and we have worked extensively with NVIDIA Tesla S1070, NVIDIA Tesla C2070 and Kepler GPUs. We have experience in parallelizing large scientific applications like cosmological simulations on GPUs.
We have a recent paper in the International Conference on Supercomputing (ICS 2013) where we present a runtime system to perform GPU specific optimizations at application run time.