Announcement

Collapse
No announcement yet.

AMD Ryzen 9 5900X On Windows 10 vs. Ubuntu Linux Performance: Windows Looks Surprisingly Good This Time

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • qarium
    replied
    Originally posted by bridgman View Post
    Just FYI we do already factor the influencer/decision-maker aspect of consumer Linux users into R&D spending. That was the rationale for restarting open source graphics driver development back in 2007, and for maintaining over 2x the R&D spend on consumer Linux relative to measurable market share.

    The limit to that argument, however, is the fact that most influencers are also aware of the quality of Windows support, so the idea that "people recommend a Windows system based entirely or substantially on the quality of Linux support" is not the whole story either.
    well honestly that just sounds much better than the market share argument alone.

    well yes every argument has its range and its limits.

    Leave a comment:


  • bridgman
    replied
    Just FYI we do already factor the influencer/decision-maker aspect of consumer Linux users into R&D spending. That was the rationale for restarting open source graphics driver development back in 2007, and for maintaining over 2x the R&D spend on consumer Linux relative to measurable market share.

    The limit to that argument, however, is the fact that most influencers are also aware of the quality of Windows support, so the idea that "people recommend a Windows system based entirely or substantially on the quality of Linux support" is not the whole story either.
    Last edited by bridgman; 12 December 2020, 07:23 PM.

    Leave a comment:


  • qarium
    replied
    Originally posted by Weasel View Post
    Ok, but companies' end goal is profit. It may not be the only one, but it's always one of the end goals.
    if money is the goal why big companies have the money to do "prioritization" on women, blacks, jews, arabs, indian, latinos, and so one and so one

    if money is the goal they just go with asian and white males only because they are more productive and cheaper. means the so called discriminated workforce.

    Originally posted by Weasel View Post
    It's not bullshit cause it's fact? He's just the messenger, he doesn't dictate where the priorities should go.
    well if i would be the manager i would not count market share as messure to invest money
    instead i would do a science study on how the people who buy windows get to know how they choose hardware and if they choose hardware because a nerd/geek tell them to buy linux compatible hardware then you can say these customers maybe buy windows hardware but they are influenced by linux people so it is a save bet that you can just move the money to linux instead of windows. because the better the linux support is the more nerd/geek people will support them in influencing the stupid sheep windows customers.

    thats science... and not the bullshit they do.

    Leave a comment:


  • Weasel
    replied
    Originally posted by Qaridarium View Post
    what is the difference between prioritization and discrimination?
    at universities all around the world white and asian men are discriminated
    because they do prioritisation on minorities like jews, blacks , latinos,
    hell they even do prioritization on majorities like women (51% of all humans are women)
    in my point of view "prioritization" is the same as "discrimination"
    Ok, but companies' end goal is profit. It may not be the only one, but it's always one of the end goals.

    Originally posted by Qaridarium View Post
    and this alone proof that this: "the testing priorities are probably closer to Enterprise Linux -> Windows -> Consumer Linux." is complete bullshit.
    It's not bullshit cause it's fact? He's just the messenger, he doesn't dictate where the priorities should go.

    Leave a comment:


  • baryluk
    replied
    So what is going on with Craft and DeCapo ? Are these results consistent across many runs? Were same JVM versions and compiler options were used? What compiler was used for the DeCapo on Windows vs Linux? How long were these tests done? I.e. 5 second test is not the same as 5 minutes test. First one would be very sensitive to noise, JIT and GC hickups, etc.

    I don't have access to Windows to test these.

    Leave a comment:


  • marios
    replied
    Something is amiss...

    The results for browser benchmarks are expected (thanks mozilla for screwing with the only platform on which your browser is relevant).

    The other results, in which windows performs better, make no sense to me. There are only 3 logical explanations for that, since most of them do not use much I/O (and the code is mostly run in user space).

    0. Different compilers (Having 0 information about Windows and not knowing the flags in ubuntu does not help at all).
    1. Bad configuration of the cpu scaling and boost stuff.
    2. Differnces in scheduling, threads and synchrinization. This can be ruled out if windows has still better performance for single threaded runs of the benchmark.

    Leave a comment:


  • qarium
    replied
    Originally posted by bridgman View Post
    I'm not going to go into "prioritization" vs "discrimination" because the end result is pretty much the same, but we are beefing up the consumer Linux coverage.
    what is the difference between prioritization and discrimination?
    at universities all around the world white and asian men are discriminated
    because they do prioritisation on minorities like jews, blacks , latinos,
    hell they even do prioritization on majorities like women (51% of all humans are women)
    in my point of view "prioritization" is the same as "discrimination"

    Originally posted by bridgman View Post
    It really boils down to market size - I don't know the exact numbers but probably something like 60% -> 39% -> 1%. The good news is that now we have stopped losing money we can move the "sorry no coverage for you" line down further.
    i tell you something what YOU and your company never expect but it is true:

    many technical people like me who use linux and only buy linux hardware do technical support for people who only use windows and in fact i never ever let a windows user buy windows-only hardware and i only only recommend linux hardware for advice and in 99% the cases they buy exactly what i tell them. And i have no university degree or professional training degree to impress people with my competence. Means i tell them buy AMD and no Nvidia/Intel.

    and i know many people like me. the technical support for windows people is in fact done by linux people THATS A FACT.

    and this alone proof that this: "the testing priorities are probably closer to Enterprise Linux -> Windows -> Consumer Linux." is complete bullshit.

    really sometimes companies who only watch market share of windows to do prioritization of their money spend is just a sign of incompetence.

    in the past you did make a similar commend like this: people who not only use linux or only use windows but use both windows and linux do not buy the windows hardware based on the windows performance but there is a high chance that they buy linux hardware instead....

    you see you get the point in the past to. companies who only watch market share of windows to do prioritization of their money spend is just a sign of incompetence

    Leave a comment:


  • bridgman
    replied
    Originally posted by Qaridarium View Post
    thats sad for me as beeing part of "consumer linux"

    i installed fedora33 cinnamon/gnome3 on my mothers computer AMD-CPU+AMD-GPU and i had to disable GPU and go with software rendering to fix crash bugs. its a FX4000(4 core bulldozer) and HD4770 yes old hardware but still sad. (on my TR 1920X+vega64 all works perfect)

    i really do not understand this "Discrimination" because if you watch steam statistic a average linux consumer is a much better consumer than a average windows user means more cores more mhz better gpu (higher vram) and so one. this means AMD does make more money on a linux consumer than on a windows consumer. (yes because of higher windows market share this maybe does not count for amd)

    in fact linux people are the very best consumers for AMD but still there is discrimination.
    I'm not going to go into "prioritization" vs "discrimination" because the end result is pretty much the same, but we are beefing up the consumer Linux coverage.

    It really boils down to market size - I don't know the exact numbers but probably something like 60% -> 39% -> 1%. The good news is that now we have stopped losing money we can move the "sorry no coverage for you" line down further.

    Leave a comment:


  • bridgman
    replied
    Originally posted by muncrief View Post
    I'm bummed out. I was going to buy a Zen 3 CPU even though my R7 3700X is already more awesomeness than I need, but they sold out so quick I missed my chance. AMD hit the ball so far out of park this time that they just can't make 'em fast enough! But hey, after all these years in the wilderness it's great to see AMD back on top of their game again.
    If it makes you feel any better I looked at 5xxx CPUs for a while then decided to go with a 3900X instead, which arrived yesterday. I haven't had a chance to do any gaming since Duke Nukem Forever, so I figured the 3900X would be pretty much as good for my workloads.

    That said, 5xxx stock seems to be showing up in stores now and sticking around for a while.

    Leave a comment:


  • qarium
    replied
    Originally posted by bridgman View Post
    Actually no - the testing priorities are probably closer to Enterprise Linux -> Windows -> Consumer Linux.
    thats sad for me as beeing part of "consumer linux"

    i installed fedora33 cinnamon/gnome3 on my mothers computer AMD-CPU+AMD-GPU and i had to disable GPU and go with software rendering to fix crash bugs. its a FX4000(4 core bulldozer) and HD4770 yes old hardware but still sad. (on my TR 1920X+vega64 all works perfect)

    i really do not understand this "Discrimination" because if you watch steam statistic a average linux consumer is a much better consumer than a average windows user means more cores more mhz better gpu (higher vram) and so one. this means AMD does make more money on a linux consumer than on a windows consumer. (yes because of higher windows market share this maybe does not count for amd)

    in fact linux people are the very best consumers for AMD but still there is discrimination.

    Leave a comment:

Working...
X