Matrix inversion in OpenCL

Question

I am trying to accelerate some computations using OpenCL and part of the algorithm consists of inverting a matrix. Is there any open-source library or freely available code to c

Answer 1

The original question (now 7 years old) actually was solved 4 years later in a paper describing matrix inversion in CUDA based on Gauss-Jordan. It attempts to distribute the calculations across different threads, and gives detailed performance indications for matrices up to 2048 in size.

While not OpenCL, the general ideas will translate from CUDA quite easily.