Search:

Type: Posts; User: brent; Keyword(s):

Page 1 of 5 1 2 3 4

Search: Search took 0.23 seconds.

  1. Replies
    71
    Views
    7,650

    mpv also supports vaapi these days, and much...

    mpv also supports vaapi these days, and much better than the old mplayer patchset ever did.



    It might be interesting to also test high FPS files. Until some time ago Radeon UVD had performance...
  2. Replies
    71
    Views
    7,650

    It is much faster compared to an E-350. There is...

    It is much faster compared to an E-350. There is definitely something very wrong with these benchmarks! See basically all the other Kabini benchmarks, which can be found on the web. The newer Jaguar...
  3. Replies
    71
    Views
    7,650

    Oh come on, it's not *that* bad. Jaguar is miles...

    Oh come on, it's not *that* bad. Jaguar is miles ahead of Bobcat. +20% IPC, much better SIMD performance and feature set, AES-NI, et cetera. On top of that it runs at significantly higher clock...
  4. I'd rather like it if they actually fixed the...

    I'd rather like it if they actually fixed the existing drivers. They are quite buggy... rtl8192ce in my netbook? Only works good without connection dropouts with power management disabled. rtl8192cu...
  5. 3.15 will include LRU policy based eviction which...

    3.15 will include LRU policy based eviction which was implemented by Marek earlier. Furthermore there are patches for two-sided allocation (top down and bottom up depending on buffer type and/or...
  6. What about IPv6 support? It's pretty laughable...

    What about IPv6 support? It's pretty laughable that this isn't a priority, but silly features like super-fast operation are. systemd-network needs to support SLAAC and DHCPv6 configurations and...
  7. Replies
    33
    Views
    4,899

    Is there any specific reason why? Sure I'd love...

    Is there any specific reason why? Sure I'd love it, but there wouldn't be any time left to fix bugs/regressions.
  8. Replies
    23
    Views
    2,615

    It's likely disabled in many implementations for...

    It's likely disabled in many implementations for historic patents reasons, unfortunately. As you noticed, the patent expired (many years ago). Arithmetic coding is part of the JPEG standard. Toggling...
  9. Replies
    23
    Views
    2,615

    It is as efficient as JPEG and might be a tad...

    It is as efficient as JPEG and might be a tad better, but it's not much better than JPEG, as Google wants you to believe. :) Alpha channel is a pretty interesting feature, I agree. It isn't what WebP...
  10. Replies
    23
    Views
    2,615

    Simple: most JPEG encoders are pretty bad,...

    Simple: most JPEG encoders are pretty bad, because they've hardly seen any improvements in the last 10-15 years; on the other hand media compression research hasn't stood still, and we can easily...
  11. Replies
    23
    Views
    2,615

    I sure hope it won't get any widescale adoption....

    I sure hope it won't get any widescale adoption. The only actual advantage of WebP is the alpha channel. On the other hand, WebP is a lot more computationally intensive to decode and doesn't support...
  12. It's pretty naive to assume that a more...

    It's pretty naive to assume that a more restrictive license will make people open up their code. There are tons of JavaScript frameworks. In the best case, developers that use ease.js will migrate to...
  13. Replies
    18
    Views
    1,975

    Note that Eric Anholt and others have very...

    Note that Eric Anholt and others have very recently put a lot of work into improving glamor greatly, e.g.:

    http://cgit.freedesktop.org/~anholt/xserver/log/?h=glamor-server

    new glamor...
  14. Meanwhile, Nvidia has been shipping a stable...

    Meanwhile, Nvidia has been shipping a stable VDPAU implementation for many years, and the Mesa VDPAU implementation for Radeon and Nouveau graphics is shaping up nicely as well. Why does Google only...
  15. An additional, important fact is that deciding to...

    An additional, important fact is that deciding to not make VDPAU available to a wide audience will not at all help with getting the remaining VDPAU issues fixed.
  16. VCE at least uses a standard API, so I'd say yes....

    VCE at least uses a standard API, so I'd say yes. It's not a proprietary single-vendor interface like VA-API, so a backend would possibly work on more hardware.
  17. Sorry, but it's not that simple, unfortunately....

    Sorry, but it's not that simple, unfortunately. They are not a simple subset, but very different and complex beasts altogether. E.g. GPUs inherently use predicated SIMD (and/or VLIW like in in R600)...
  18. The IR is fine, but the optimization passes...

    The IR is fine, but the optimization passes apparently aren't very suitable for GPUs.
  19. However, LLVM was designed for general purpose...

    However, LLVM was designed for general purpose CPUs. GPUs have very different architectures. So there are more than "a little" modifications required.
  20. I can sort of understand this, even though I...

    I can sort of understand this, even though I don't know LLVM well. Just observing the public development is quite interesting.

    AMD has been working on LLVM backends for r600 and radeonsi hardware...
Results 1 to 20 of 100
Page 1 of 5 1 2 3 4