Announcement

Collapse
No announcement yet.

Benchmarks Of The Gentoo-based Sabayon

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

  • n0nsense
    replied
    Originally posted by tuxdriver View Post
    But can you beat this one? :



    Now that's what I call a *lean* distro

    I haven't seen any other distro that runs so fast with the Xorg 'vesa' driver.
    Try Gentoo Live CD with docahce boot option
    You'll see what is really fast

    Anyway, you can have same configuration with Gentoo and even with ubuntu.
    Gentoo with XFCE or some other minimal environment should be about same size.
    But Gentoo never meant to be smallest distro (at least disk space).
    You can definitely achieve this goal on Gentoo since you can optimize you system toward size by making GCC to produce smaller binaries at cost of compilation time. and then just remove anything that is not necessary to boot.
    It is all about customizations and optimizations
    For example if you want (or don't want) support for something on your system, it is very simple.
    For example add -cups to make.conf use flags, scan package.use for packages that explicitly require cups (comment them). and just "emerge -vuND world" no more cups on your system
    (I have edited the example to replace mono with cups)

    Leave a comment:


  • Smorg
    replied
    Last stage 3 size I checked:
    Code:
    -rw-r--r--   1 root root 134M Nov 12 05:29 stage3-amd64-20091112.tar.bz2
    -rw-r--r--   1 root root  35M Nov 18 19:45 portage-latest.tar.bz2
    Gentoo has a hard dependency on python which will make it a bit larger. Apparently turbopup is around ~100MB; DSL is 50, if you're going by disk space.

    People who use words like "bloat", think top is giving them an accurate picture of memory usage, and think that the amount of disk space taken up by a base install is somehow connected with performance really don't understand how memory allocation and shared libs work.

    This distro will suffer the same problems that any distro not tuned for specific hardware will suffer.

    Leave a comment:


  • clavko
    replied
    Originally posted by leidola View Post
    This seems rather slow. I also compiled Qt on my eee and if I recall right it was faster, maybe an hour or two. Did you set the pch use flag?
    No, the pch flag hasn't been used, perhaps I should've turned it on.
    I glazed briefly through pch description and saw the disk expense remark
    but now I'm inclined to think that would be only during compilation.

    On the other side... Perhaps there are other differences, I'm currently
    on x86, so I can't use additional registers for compiling (so I heard)
    and since compiling is somewhat different than the usual computer usage,
    I'm not sure whether Atom's architeture has some advantages over K8 in
    that sense, but the instruction sets are comparable.

    Anyways, I'm definitely switching to x64 in a week or so, I'll report
    anything worth reporting. Thanks for the tip!

    Leave a comment:


  • tuxdriver
    replied
    Bare minimum services running are just as easily done on other distros, openSUSE for example.
    But can you beat this one? :



    Now that's what I call a *lean* distro

    I haven't seen any other distro that runs so fast with the Xorg 'vesa' driver.
    Last edited by tuxdriver; 05 January 2010, 11:36 PM.

    Leave a comment:


  • korpenkraxar
    replied
    Back in the day I used Gentoo because it was great fun and interesting. Now I use Arch because it is great fun and interesting. I've used many many distros over the years, most of which were great fun and interesting. For speed I got myself a new cheap CPU and harddrive upgrade every now and then.

    I use Linux because it is great fun and interesting.
    Last edited by korpenkraxar; 05 January 2010, 07:57 PM.

    Leave a comment:


  • leidola
    replied
    Originally posted by clavko View Post
    Just yesterday it took me about 2-3 hours to compile Qt 4.6.1
    This seems rather slow. I also compiled Qt on my eee and if I recall right it was faster, maybe an hour or two. Did you set the pch use flag?

    Leave a comment:


  • deanjo
    replied
    Originally posted by yogi_berra View Post
    But which leaves the user in better shape to deal with issues that arise? The extensive documentation of Gentoo or the hand-holding/canned setup of Novell?
    Novell's documentation is nothing to sneeze at either. Documentation in fact has always been one of SuSE's strong points. A hand holding approach works for most people but the documentation is also there for the more demanding, your just not forced into having to use it.

    Leave a comment:


  • yogi_berra
    replied
    Originally posted by deanjo View Post
    Compared to having to read extensive documentation to setup a Gentoo system most definitely.
    But which leaves the user in better shape to deal with issues that arise? The extensive documentation of Gentoo or the hand-holding/canned setup of Novell?

    Leave a comment:


  • Smorg
    replied
    Benchmarking the differences in the aggregate effects of the default configuration of a large collection of software (a distribution) against those of essentially the same collection whose defaults vary to unknown degrees, and concluding that therefore distro x is faster/better than distro y is unscientific and misleading. There are too many variables. If you want to determine causality you have to vary just one thing at a time. Any of the benchmarks could be affected by a single change thats possible within any distro.

    Comparing Gentoo as a whole with other distros would be utterly pointless. In the case of Gentoo, choice is valued over defaults, and defaults are only made to be sensible and implicit in the event a default is even necessary. Any performance advantage noticed in any distro made of the same potential software as Gentoo should also be achievable in Gentoo itself depending upon what you have installed and how you have it configured.

    If you care about meaningless things like system-wide benchmarks and "defaults", Gentoo is absolutely the system you should be using. The need to compile things is not at all a big deal, especially on modern hardware. After the initial installation, the incremental version changes should not take unbearably long, and even if they do, its worth it for the power you get. Rarely do you actually need instant updates... you can still work while your system is updating.
    Last edited by Smorg; 05 January 2010, 04:09 PM.

    Leave a comment:


  • kraftman
    replied
    Originally posted by n0nsense View Post
    There is no such thing as "generic" Gentoo.
    It is always "your Gentoo"
    Most close to "Generic Gentoo" is Live CD/DVD.
    There is also Sabayon But in this case, you have to
    • Install Sabayon
    • Edit make.conf
    • Rebuild the system


    And yet, it isn't generic Gentoo.
    You're right. I meant something like you already pointed

    @Clavko

    I agree. That's why I'm trying to say. For benchmarking to have some sense
    one should have as much parameters as it can under control. Gentoo has this
    kind of advantage to be a perfect customizable benchmark environment.
    Now, I know it's time consuming, that's why I don't expect Phoronix to
    jump at it. However... mmm... it sure as hell would be nice
    Yes, I completely agree with you. It would be nice to see Phoronix doing such thing

    Leave a comment:

Working...
X