Page 2 of 2 FirstFirst 12
Results 11 to 20 of 20

Thread: Handbrake 0.9.9 Supports OpenCL Offloading

  1. #11

    Default

    I hope 1.0 will support WebM with VP9 and Opus.

  2. #12
    Join Date
    Jan 2010
    Posts
    363

    Default

    Combined with accelerated video decoding, it makes sense to offload more stuff to the GPU. This way all CPU cycles can be devoted to encoding, and the bus traffic is minimal. Handbrake would only need to push the compressed source video data to the GPU and would read back cropped and scaled frames, and encode these. I'm sure that is what it is doing...? The release notes aren't very clear about it.
    Last edited by brent; 05-21-2013 at 10:25 AM.

  3. #13
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,583

    Default

    Quote Originally Posted by JohnAStebbins View Post
    People are not reading the announcement very carefully.
    1. HandBrake OpenCL is Beta
    2. It's not in the 0.9.9 release

    As a rule, HandBrake doesn't release major features till they work on *all* platforms we claim to support (windows, osx, linux). So when it's ready, it should be working on all. Intel QuickSync is also in the pipeline.

    Here's the actual announcements if you care to get it from the source http://handbrake.fr/news.php
    Interesting you would say openCL 1.2 and NVidia in the notes. NVidia as of right now only supports 1.1. What 1.2 features are being used (if any).

  4. #14
    Join Date
    Sep 2008
    Posts
    40

    Default

    Quote Originally Posted by pingufunkybeat View Post
    It would be nice to see some benchmarks. It seems that the supported operations are cropping and scaling.

    Given that the PCI transfer is extremely expensive, and scaling and cropping rather cheap operations, I'm surprised that this is worthwhile.

    On the other hand, it's probably just a first step, with more complex operations planned in the future.
    It actually makes sense, especially for AMD whose future APU architectures will have shared CPU/GPU memory addressing. Currently, although they share physical memory, there's still a copy involved.

    Like you said, it makes less sense for discrete graphics cards.

  5. #15
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,583

    Default

    Quote Originally Posted by pingufunkybeat View Post
    It would be nice to see some benchmarks. It seems that the supported operations are cropping and scaling.

    Given that the PCI transfer is extremely expensive, and scaling and cropping rather cheap operations, I'm surprised that this is worthwhile.

    On the other hand, it's probably just a first step, with more complex operations planned in the future.
    Well, just tried a couple of encodes on my 8350/titan system. Didn't make a whole lot of difference on a quality based encode, resized from 1080 to 720 with cropping. With openCL 20m32s, without openCL 20m3 seconds. It was using the Titan but barely, gpu load never exceeded 9%. Now QuickSync accelerated on the other hand......6m5s
    Last edited by deanjo; 05-21-2013 at 11:32 PM.

  6. #16

    Default

    Quote Originally Posted by deanjo View Post
    With openCL 20m32s, without openCL 20m3 seconds. It was using the Titan but barely, gpu load never exceeded 9%.
    Maybe because OpenCL is always second class citizen on nVidia hardware? Example: http://www.luxrender.net/forum/viewt...hp?f=17&t=9646

  7. #17
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,583

    Default

    Quote Originally Posted by RussianNeuroMancer View Post
    Maybe because OpenCL is always second class citizen on nVidia hardware? Example: http://www.luxrender.net/forum/viewt...hp?f=17&t=9646
    Not really, results are just as unspectacular on AMD cards in handbrake.

  8. #18
    Join Date
    Oct 2008
    Posts
    3,098

    Default

    Quote Originally Posted by deanjo View Post
    Not really, results are just as unspectacular on AMD cards in handbrake.
    I suspect this is really meant for AMD's upcoming HSA architecture that integrates the GPU more closely in the CPU.

    I think they're supposed to come out later this year, and this will give AMD a benchmark to point to which shows them beating out Intel.

  9. #19
    Join Date
    Jul 2008
    Location
    Germany
    Posts
    650

    Default

    If i remember correctly the OpenCL Part is only for the lookahead and that is not really important, but the DXVA decoder brings a benefit.

  10. #20
    Join Date
    Jun 2013
    Posts
    1

    Default

    Quote Originally Posted by plonoma View Post
    You have to transfer the video to the graphics card anyway because of the connection to the display.

    The most optimal way would be to load the video into the graphics card memory and decode, composite and post-process on the graphics card. The graphics card is much better suited for graphics calculations and it frees the CPU for other tasks.
    AppGeeker handle this job well, tool.
    http://www.ilikemall.com/how-to/conv...o-mov-mac.html

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •