gpu acceleration into FFmpeg
- Thread starter Okedokey
- Start date
S.T.A.R.S.
banned
I think that Fmpeg today doesn't support any means of GPU encoding outside of the libx264.
x264 supports openCL for some lookahead operations I think,but there are also some encoders that utilize the GPU,but that does not mean they are well optimized lol even though encoding time may be faster and they might be worse than Vanilla x264 (so I read).But I never really tried that library so feel free to give it a try yourself.
The thing is that most of them depend on the CPU only and very rare have an option to use the GPU.I don't know if the time taken to copy the raw video to the GPU encode and then read back the compressed data would really mean there was much benefit from a GPU in the first place lol.
With the playback of some video you are forced to copy the data to the GPU anyway so it's obvisuly better to copy the smaller compressed source stream.
That is because of data dependencies because the GPU must use an iterative motion search which performs more total work than the CPU would do so this is not work efficient or power efficient if you ask me lol.
But if there are spare GPU cycles to spare then it can often to speed up the encode.Output quality when openCL lookahead is enabled is often very slightly worse in quality than the CPU quality because of the same data dependencies.
Here are some more informations about openCL:
http://git.videolan.org/?p=x264.git;a=commit;h=3a5f6c0aeacfcb21e7853ab4879f23ec8ae5e042
Cheers to you to!
x264 supports openCL for some lookahead operations I think,but there are also some encoders that utilize the GPU,but that does not mean they are well optimized lol even though encoding time may be faster and they might be worse than Vanilla x264 (so I read).But I never really tried that library so feel free to give it a try yourself.
The thing is that most of them depend on the CPU only and very rare have an option to use the GPU.I don't know if the time taken to copy the raw video to the GPU encode and then read back the compressed data would really mean there was much benefit from a GPU in the first place lol.
With the playback of some video you are forced to copy the data to the GPU anyway so it's obvisuly better to copy the smaller compressed source stream.
That is because of data dependencies because the GPU must use an iterative motion search which performs more total work than the CPU would do so this is not work efficient or power efficient if you ask me lol.
But if there are spare GPU cycles to spare then it can often to speed up the encode.Output quality when openCL lookahead is enabled is often very slightly worse in quality than the CPU quality because of the same data dependencies.
Here are some more informations about openCL:
http://git.videolan.org/?p=x264.git;a=commit;h=3a5f6c0aeacfcb21e7853ab4879f23ec8ae5e042
Cheers to you to!