Hello,
I have been using the NVIDIA hardware accelerated video encoding and decoding. The objective is to process as many videos as possible in parallel. As per my understanding, this process does not use the main cores of the GPU. Rather it uses separate (limited number of) NVENC chips. Therefore, a huge part of my GPU is not being used during the process.
I do understand that not all steps of encoding/decoding are fully suitable for parallel computation. However, are there any other implementations of encoders/decoders(H264/H265) that can utilize the large parallel compute resource (main cores) of GPU and result in increased processing speeds as well as better quality?