NVENC using PBO pointer

Accelerated Computing GPU-Accelerated Libraries
1

I am trying to use the NVENC H264 hardware encoder (using Quadro K4000 card) to encode video where each frame is the result of several OpenGL pre-processing functions.
The processed frame resides in YUV format in a PBO object.
What is the most efficient way to pass this PBO pointer (or pointers into each YUV plane) to the encoder as input data?

Thanks,
Shachar

2

did you ever figure out a way to do this?

3

Also interested, in my case it’s a FBO but I guess it is similar. I’m guessing CUDA interop? But some guidelines would be nice :)