search

Chen-tao / webm

Automatically exported from code.google.com/p/webm
0 stars 0 forks source link

Could the vpx be parallelized on GPU for more speedup? #550

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago 
There are four moduls for vpx mainly.
=============================
Inverse entropy code
Inverse transform (DCT/WHT)
Intra and inter prediction;
loopfilter.
=============================
They're all run on cpu.
And for 1080p, the decoded time for one frame is about 13ms by
single threaded.If use ffmpeg by using vp8 decoder, the time is about 3.9ms by 
multi threaded.
So my question is that
Could the vpx be parallelized on GPU for more speedup?

Original issue reported on code.google.com by lemontre...@gmail.com on 22 Feb 2013 at 2:25

GoogleCodeExporter commented 8 years ago 
It's not on the current roadmap. There has been some promising work on this 
though, see:

[1]: https://github.com/awatry/libvpx.opencl

Original comment by jkoles...@google.com on 26 Mar 2013 at 4:16