I’m a self-taught weekend warrior with OpenGL. I have a complex film scanning application that processes sequential movie film frames. Upload, render and download are are currently performed in a single, linear thread.
I’m trying to port my code over to multithreaded OpenGL code to take advantage of Fermi dual copy engines. I have the sample code from the Nvidia book, but the code is very encapsulated and is a little cryptic for me to use for learning the synchronization calls.
I have two beginning questions:
-
Will the upload/download copy engines work if I don’t use Pixel Buffer Objects for upload and download? I wanted to start without that, and then add it once I have the copy engine multithreaded code working.
-
Is there any sample code available for simultaneous texture upload, render and download (using multithreaded OpenGl and Copy engines) that uses the native OpenGl C calls, as opposed to the C++ class used in the OpenGL official sample code?
Thanks.