Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

LAwLz

Member
  • Content Count

    12,823
  • Joined

  • Last visited


Reputation Activity

  1. Like
    LAwLz got a reaction from Syntaxvgm for a status update, Benchmarks of the current Dav1d (AV1 software decoder) are out . Summary: Dav1d scale   
    Benchmarks of the current Dav1d (AV1 software decoder) are out.
     
    Summary:
    Dav1d scales very well with more cores and threads. They tried out a 32-core Epyc processor and it was still showing signs of increased performance for each additional thread when decoding a 1080p clip.
     
    Performance is essentially cut in half going from 8 bit color depth to 10 or 12 bit color depth.
     
    Firefox does not have the necessary components for using the assembly code and instead relies 100% on dav1d written in C. The C version is roughly 50% slower than the assembly version.
     
    A quad core Haswell processor is getting around 30-50 FPS decoding 4K video. For 1080p the same processor gets well over 100 FPS (in some cases, close to 200 FPS) doing decoding fully in software.
     
    Dav1d seems to be around 2 times as fast as the reference aomdec (AOMedia's decoder) in general, although in some clips it is up to 5 times as fast.
     
    On mobile, the Snapdragon 835 gets around 50 FPS playing a 1080p clip.
    The Apple A12X gets something like 120 FPS.
     
     
    Their next focus is improving 10 bit support, as well as SSSE3 processors and ARM64 processors.
  2. Like
    LAwLz got a reaction from Syntaxvgm for a status update, Benchmarks of the current Dav1d (AV1 software decoder) are out . Summary: Dav1d scale   
    Benchmarks of the current Dav1d (AV1 software decoder) are out.
     
    Summary:
    Dav1d scales very well with more cores and threads. They tried out a 32-core Epyc processor and it was still showing signs of increased performance for each additional thread when decoding a 1080p clip.
     
    Performance is essentially cut in half going from 8 bit color depth to 10 or 12 bit color depth.
     
    Firefox does not have the necessary components for using the assembly code and instead relies 100% on dav1d written in C. The C version is roughly 50% slower than the assembly version.
     
    A quad core Haswell processor is getting around 30-50 FPS decoding 4K video. For 1080p the same processor gets well over 100 FPS (in some cases, close to 200 FPS) doing decoding fully in software.
     
    Dav1d seems to be around 2 times as fast as the reference aomdec (AOMedia's decoder) in general, although in some clips it is up to 5 times as fast.
     
    On mobile, the Snapdragon 835 gets around 50 FPS playing a 1080p clip.
    The Apple A12X gets something like 120 FPS.
     
     
    Their next focus is improving 10 bit support, as well as SSSE3 processors and ARM64 processors.
  3. Like
    LAwLz got a reaction from Syntaxvgm for a status update, Benchmarks of the current Dav1d (AV1 software decoder) are out . Summary: Dav1d scale   
    Benchmarks of the current Dav1d (AV1 software decoder) are out.
     
    Summary:
    Dav1d scales very well with more cores and threads. They tried out a 32-core Epyc processor and it was still showing signs of increased performance for each additional thread when decoding a 1080p clip.
     
    Performance is essentially cut in half going from 8 bit color depth to 10 or 12 bit color depth.
     
    Firefox does not have the necessary components for using the assembly code and instead relies 100% on dav1d written in C. The C version is roughly 50% slower than the assembly version.
     
    A quad core Haswell processor is getting around 30-50 FPS decoding 4K video. For 1080p the same processor gets well over 100 FPS (in some cases, close to 200 FPS) doing decoding fully in software.
     
    Dav1d seems to be around 2 times as fast as the reference aomdec (AOMedia's decoder) in general, although in some clips it is up to 5 times as fast.
     
    On mobile, the Snapdragon 835 gets around 50 FPS playing a 1080p clip.
    The Apple A12X gets something like 120 FPS.
     
     
    Their next focus is improving 10 bit support, as well as SSSE3 processors and ARM64 processors.
  4. Like
    LAwLz got a reaction from Syntaxvgm for a status update, Benchmarks of the current Dav1d (AV1 software decoder) are out . Summary: Dav1d scale   
    Benchmarks of the current Dav1d (AV1 software decoder) are out.
     
    Summary:
    Dav1d scales very well with more cores and threads. They tried out a 32-core Epyc processor and it was still showing signs of increased performance for each additional thread when decoding a 1080p clip.
     
    Performance is essentially cut in half going from 8 bit color depth to 10 or 12 bit color depth.
     
    Firefox does not have the necessary components for using the assembly code and instead relies 100% on dav1d written in C. The C version is roughly 50% slower than the assembly version.
     
    A quad core Haswell processor is getting around 30-50 FPS decoding 4K video. For 1080p the same processor gets well over 100 FPS (in some cases, close to 200 FPS) doing decoding fully in software.
     
    Dav1d seems to be around 2 times as fast as the reference aomdec (AOMedia's decoder) in general, although in some clips it is up to 5 times as fast.
     
    On mobile, the Snapdragon 835 gets around 50 FPS playing a 1080p clip.
    The Apple A12X gets something like 120 FPS.
     
     
    Their next focus is improving 10 bit support, as well as SSSE3 processors and ARM64 processors.
  5. Like
    LAwLz got a reaction from Syntaxvgm for a status update, Benchmarks of the current Dav1d (AV1 software decoder) are out . Summary: Dav1d scale   
    Benchmarks of the current Dav1d (AV1 software decoder) are out.
     
    Summary:
    Dav1d scales very well with more cores and threads. They tried out a 32-core Epyc processor and it was still showing signs of increased performance for each additional thread when decoding a 1080p clip.
     
    Performance is essentially cut in half going from 8 bit color depth to 10 or 12 bit color depth.
     
    Firefox does not have the necessary components for using the assembly code and instead relies 100% on dav1d written in C. The C version is roughly 50% slower than the assembly version.
     
    A quad core Haswell processor is getting around 30-50 FPS decoding 4K video. For 1080p the same processor gets well over 100 FPS (in some cases, close to 200 FPS) doing decoding fully in software.
     
    Dav1d seems to be around 2 times as fast as the reference aomdec (AOMedia's decoder) in general, although in some clips it is up to 5 times as fast.
     
    On mobile, the Snapdragon 835 gets around 50 FPS playing a 1080p clip.
    The Apple A12X gets something like 120 FPS.
     
     
    Their next focus is improving 10 bit support, as well as SSSE3 processors and ARM64 processors.
  6. Agree
    LAwLz got a reaction from Techstorm970 for a status update, Don't like YouTube Rewind? Wait for Google Zeitgeist 2018 instead. It's kind of like   
    Don't like YouTube Rewind? Wait for Google Zeitgeist 2018 instead.
    It's kind of like "Youtube rewird", but for Google search, and actually just includes relevant things. It's because it's based on what people searched for throughout the year, rather than some facade Youtube tries to put up.
     
    2010 2011 2012
    2013 2014 2015
    2016 2017
  7. Informative
    LAwLz got a reaction from TVwazhere for a status update, Analysis of the Apple USB-C headphone adapter, and also a comparison between it and G   
    Analysis of the Apple USB-C headphone adapter, and also a comparison between it and Google's adapters.
    Spoiler: The Apple one is actually pretty good, and outclasses the Google one despite being cheaper.
     
    https://www.audiosciencereview.com/forum/index.php?threads/review-apple-vs-google-usb-c-headphone-adapters.5541/
  8. Informative
    LAwLz got a reaction from TVwazhere for a status update, Analysis of the Apple USB-C headphone adapter, and also a comparison between it and G   
    Analysis of the Apple USB-C headphone adapter, and also a comparison between it and Google's adapters.
    Spoiler: The Apple one is actually pretty good, and outclasses the Google one despite being cheaper.
     
    https://www.audiosciencereview.com/forum/index.php?threads/review-apple-vs-google-usb-c-headphone-adapters.5541/
  9. Informative
    LAwLz got a reaction from TVwazhere for a status update, Analysis of the Apple USB-C headphone adapter, and also a comparison between it and G   
    Analysis of the Apple USB-C headphone adapter, and also a comparison between it and Google's adapters.
    Spoiler: The Apple one is actually pretty good, and outclasses the Google one despite being cheaper.
     
    https://www.audiosciencereview.com/forum/index.php?threads/review-apple-vs-google-usb-c-headphone-adapters.5541/
  10. Agree
    LAwLz got a reaction from Techstorm970 for a status update, Don't like YouTube Rewind? Wait for Google Zeitgeist 2018 instead. It's kind of like   
    Don't like YouTube Rewind? Wait for Google Zeitgeist 2018 instead.
    It's kind of like "Youtube rewird", but for Google search, and actually just includes relevant things. It's because it's based on what people searched for throughout the year, rather than some facade Youtube tries to put up.
     
    2010 2011 2012
    2013 2014 2015
    2016 2017
  11. Agree
    LAwLz got a reaction from Syntaxvgm for a status update, Australia is still pushing their anti-encryption bill. https://techcrunch.com/2018/12   
    Australia is still pushing their anti-encryption bill.
    https://techcrunch.com/2018/12/05/australia-rushes-its-dangerous-anti-encryption-bill-into-parliament/
     
    EFF has a good article on why the bill is bad, that I highly recommend reading:
    https://www.eff.org/deeplinks/2018/09/australian-government-ignores-experts-advancing-its-anti-encryption-bill
  12. Agree
    LAwLz got a reaction from Syntaxvgm for a status update, Australia is still pushing their anti-encryption bill. https://techcrunch.com/2018/12   
    Australia is still pushing their anti-encryption bill.
    https://techcrunch.com/2018/12/05/australia-rushes-its-dangerous-anti-encryption-bill-into-parliament/
     
    EFF has a good article on why the bill is bad, that I highly recommend reading:
    https://www.eff.org/deeplinks/2018/09/australian-government-ignores-experts-advancing-its-anti-encryption-bill
  13. Agree
    LAwLz got a reaction from Syntaxvgm for a status update, Australia is still pushing their anti-encryption bill. https://techcrunch.com/2018/12   
    Australia is still pushing their anti-encryption bill.
    https://techcrunch.com/2018/12/05/australia-rushes-its-dangerous-anti-encryption-bill-into-parliament/
     
    EFF has a good article on why the bill is bad, that I highly recommend reading:
    https://www.eff.org/deeplinks/2018/09/australian-government-ignores-experts-advancing-its-anti-encryption-bill
  14. Agree
    LAwLz got a reaction from Cyberspirit for a status update, Just checked Windows update and I still have not received the October update. Not sur   
    Just checked Windows update and I still have not received the October update. Not sure if I should feel blessed or cursed.
    My guess is that I have something Microsoft has flagged as incompatible with the update, like they did to some Intel GPU drivers.
     
    My question is, why not display the reason for not getting the update? They clearly have a tool which scans the system for things and if it detects certain things the update doesn't get downloaded. All they have to do is print a snippet of text with the conflict they found. But nope, instead users are kept in the dark and have to guess why they may not be getting updates.
  15. Agree
    LAwLz got a reaction from Cyberspirit for a status update, Just checked Windows update and I still have not received the October update. Not sur   
    Just checked Windows update and I still have not received the October update. Not sure if I should feel blessed or cursed.
    My guess is that I have something Microsoft has flagged as incompatible with the update, like they did to some Intel GPU drivers.
     
    My question is, why not display the reason for not getting the update? They clearly have a tool which scans the system for things and if it detects certain things the update doesn't get downloaded. All they have to do is print a snippet of text with the conflict they found. But nope, instead users are kept in the dark and have to guess why they may not be getting updates.
  16. Agree
    LAwLz got a reaction from Cyberspirit for a status update, Just checked Windows update and I still have not received the October update. Not sur   
    Just checked Windows update and I still have not received the October update. Not sure if I should feel blessed or cursed.
    My guess is that I have something Microsoft has flagged as incompatible with the update, like they did to some Intel GPU drivers.
     
    My question is, why not display the reason for not getting the update? They clearly have a tool which scans the system for things and if it detects certain things the update doesn't get downloaded. All they have to do is print a snippet of text with the conflict they found. But nope, instead users are kept in the dark and have to guess why they may not be getting updates.
  17. Funny
    LAwLz got a reaction from scottyseng for a status update,   
  18. Funny
    LAwLz got a reaction from scottyseng for a status update,   
  19. Funny
    LAwLz got a reaction from scottyseng for a status update,   
  20. Funny
    LAwLz reacted to mr moose for a status update, It's funny watching kids use the "funny" button to indicate sarcastic derision. It's   
    It's funny watching kids use the "funny" button to indicate sarcastic derision.  It's like watching children in the schoolyard name call.
  21. Informative
    LAwLz got a reaction from MoonSpot for a status update, More Spectre and Meltdown vulnerabilities discovered. 7 new ones to be more precise.   
    More Spectre and Meltdown vulnerabilities discovered. 7 new ones to be more precise.
    AMD, Intel and ARM processors are vulnerable.
     
    Intel has come forth and said that some of the already existing patches stops some of these attacks though. Not sure if that's accurate and if it is, not sure if it also applies to AMD and ARM processors or if they are still vulnerable even with patches applied.
     
    Source:
    https://www.zdnet.com/article/researchers-discover-seven-new-meltdown-and-spectre-attacks/
  22. Informative
    LAwLz got a reaction from MoonSpot for a status update, More Spectre and Meltdown vulnerabilities discovered. 7 new ones to be more precise.   
    More Spectre and Meltdown vulnerabilities discovered. 7 new ones to be more precise.
    AMD, Intel and ARM processors are vulnerable.
     
    Intel has come forth and said that some of the already existing patches stops some of these attacks though. Not sure if that's accurate and if it is, not sure if it also applies to AMD and ARM processors or if they are still vulnerable even with patches applied.
     
    Source:
    https://www.zdnet.com/article/researchers-discover-seven-new-meltdown-and-spectre-attacks/
  23. Informative
    LAwLz got a reaction from MoonSpot for a status update, More Spectre and Meltdown vulnerabilities discovered. 7 new ones to be more precise.   
    More Spectre and Meltdown vulnerabilities discovered. 7 new ones to be more precise.
    AMD, Intel and ARM processors are vulnerable.
     
    Intel has come forth and said that some of the already existing patches stops some of these attacks though. Not sure if that's accurate and if it is, not sure if it also applies to AMD and ARM processors or if they are still vulnerable even with patches applied.
     
    Source:
    https://www.zdnet.com/article/researchers-discover-seven-new-meltdown-and-spectre-attacks/
  24. Funny
    LAwLz got a reaction from TopHatProductions115 for a status update, It's quite amazing how many things Microsoft appears to be breaking with update 1809.   
    It's quite amazing how many things Microsoft appears to be breaking with update 1809.
    First people discovered that the update was deleting files on their computers and the update was pulled.
     
    Then other users noticed that the function for detecting duplicate files was in some instances broken, and would just overwrite files without prompting the user.
     
    And apparently the update currently (I don't think Microsoft has re-released it yet but I could be wrong) has a quite big performance impact according to this article.
     
     
    I remember when I voiced concerns for removing control of updates from users, and some users on this forum told me, over and over, that Microsoft were not going to release broken updates because they were tested thoroughly before being rolled out and the telemetry data they collected...
  25. Funny
    LAwLz got a reaction from TopHatProductions115 for a status update, It's quite amazing how many things Microsoft appears to be breaking with update 1809.   
    It's quite amazing how many things Microsoft appears to be breaking with update 1809.
    First people discovered that the update was deleting files on their computers and the update was pulled.
     
    Then other users noticed that the function for detecting duplicate files was in some instances broken, and would just overwrite files without prompting the user.
     
    And apparently the update currently (I don't think Microsoft has re-released it yet but I could be wrong) has a quite big performance impact according to this article.
     
     
    I remember when I voiced concerns for removing control of updates from users, and some users on this forum told me, over and over, that Microsoft were not going to release broken updates because they were tested thoroughly before being rolled out and the telemetry data they collected...
×