Skip to content
This repository has been archived by the owner on Jul 21, 2020. It is now read-only.

VP8 CUDA? #65

Open
Luckz opened this issue Jan 2, 2017 · 3 comments
Open

VP8 CUDA? #65

Luckz opened this issue Jan 2, 2017 · 3 comments

Comments

@Luckz
Copy link

Luckz commented Jan 2, 2017

https://github.com/Italtel-Unimi/libvpx is CUDA-accelerated libvpx, would this be useful for HW accelerated streaming? Or is NVENC ultimately the better choice?

@Maximilian-Reuter
Copy link

From what I understand, this does not encode in Hardware but vie GPGPU computing which will eat into your available graphics performance. NVENC on the other hand is entirely in hardware (depending on the exact type of codec and generation used). So ultimately NVENC seems to be the better choice except for very low fidelity games/use cases with very high cpu demands.

Also I'm not sure this discussion belongs in this repository and rather into https://github.com/jp9000/obs-studio

@Luckz
Copy link
Author

Luckz commented Jan 8, 2017

Does OBS also use VP8 then?

@Maximilian-Reuter
Copy link

Maximilian-Reuter commented Jan 8, 2017

nope it uses H264, tachyon, which is basically an older OBS with most of the options stripped out, only uses vp8 because at the beginning of Beam the WEBRTC standard wasn't really supporting h264. But this has been changed for quite a while now. This is also a reason why this repo exists.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Development

No branches or pull requests

2 participants