Not sure if this is the right place for this post(I've posted on the DirectGMA list) but here is my question:
I would like to move streamed video over IP from my network card to GPU memory with the minimum latency.
But, I don't want to restrict myself to FirePRO cards: I would like to use the upcoming 3xx series of cards.
So, assume I am writing a driver for my network card, on linux, and have full contol over the network card buffers,
how do I transfer to GPU memory with highest speed and lowest latency?
I am posting on the OpenCL group because I am planning to process the video using OpenCL.