Before there was CUDA or OpenCL people were using GPUs for computation. I am trying to find out how they did that -- because I want to press my Rasberry Pi's GPU for computing and it does not seem to have OpenCL support.
I am looking for notes on how they did this.
I've done cursory Googling but have come up empty handed -- primarily because I think the search indexes and full of CUDA and OpenCL related material.
Would appreciate any pointers on how to use GPUs for computation in the absence of CUDA and OpenCL.