[FFmpeg-devel] [PATCH v5] Improved the performance of 1 decode + N filter graphs and adaptive bitrate.

[Wang, Shaofei]

Thanks. Seems I need to cover external samples either

> -----Original Message-----
> From: ffmpeg-devel [mailto:ffmpeg-devel-bounces at ffmpeg.org] On Behalf Of
> Michael Niedermayer
> Sent: Saturday, February 16, 2019 5:22 AM
> To: FFmpeg development discussions and patches <ffmpeg-devel at ffmpeg.org>
> Subject: Re: [FFmpeg-devel] [PATCH v5] Improved the performance of 1
> decode + N filter graphs and adaptive bitrate.
> 
> On Fri, Feb 15, 2019 at 04:54:23PM -0500, Shaofei Wang wrote:
> > It enabled multiple filter graph concurrency, which bring above about
> > 4%~20% improvement in some 1:N scenarios by CPU or GPU acceleration
> >
> > Below are some test cases and comparison as reference.
> > (Hardware platform: Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz)
> > (Software: Intel iHD driver - 16.9.00100, CentOS 7)
> 
> breaks fate
> 
> make -j12 fate-filter-overlay-dvdsub-2397 V=2
> 
> 
> frame=  208 fps=0.0 q=-0.0 Lsize=      48kB time=00:00:08.04 bitrate=
> 49.0kbits/s speed=10.6x
> video:105300kB audio:1254kB subtitle:0kB other streams:0kB global
> headers:0kB muxing overhead: unknown pthread_join failed with error: No
> such process Aborted (core dumped)
> make: *** [fate-filter-overlay-dvdsub-2397] Error 134
> 
> [...]
> --
> Michael     GnuPG fingerprint:
> 9FF2128B147EF6730BADF133611EC787040B0FAB
> 
> Asymptotically faster algorithms should always be preferred if you have
> asymptotical amounts of data