we are using OpenGL 2.1 extensions to accelerate some functions for our NLE software based on Intel CPU and IPP. The very essential code works finely on all ATI and Nvidia cards but it is amazing slow if we try to use this code on Intel HD 3000. The problem affects the copy of any video frame from GPU Memory to system memory. We have created a small benchmark for Intel engineers hoping this can help to fix the problem. We have already our h264 encoder based on new Intel media SDK (AVX and Quicksync) and I'd like to use it but we cant do it because the OpenGL iussue is pushing us to Adopt Nvidia or AMD cards dropping the Quicksync support.