What didn't work?

Obviously, even though all three 64-bit systems have matured significantly in the last 18 months, there were multiple issues with our hardware that needed to be addressed. Our first primary concern came when we were unable to install the NVIDIA graphics driver for our Fedora distribution. Granted, we were capable of recompiling the kernel and doing the test over again, but this would render our "out of the box" mentality useless. The strict guidelines that we set down for this analysis was to observe the behavior of Operating Systems out of the box - with only basic up-to-date driver sets installed.

Our largest Linux annoyance when working with these benchmarks was the lack of support for the newest on-core NVIDIA gigabit Ethernet. In this sense, our delayed Linux nForce3 250 launch has not been successful. Fedora Core 2 failed to play nicely with NVIDIA when our x86-64 base did not come precompiled with support for NVIDIA's driver set.




Click to enlarge.


This was thoroughly discouraging; no out-of-the-box NVIDIA support for the largest (or at least second largest) 64-bit operating system.

On a side note, we have to give SuSE much credit for YAST2. YAST is SuSE's install program that, while somewhat bloatish, does an excellent job of keeping software and drivers updated on a near-Windows level of simplicity (think Windows Update). It is unfortunate that our Linux benchmarks are limited to only such a small cross-section due to software availability.

All NVIDIA drivers worked fine on Windows XP 64-bit BETA, but nothing was digitally signed by Microsoft yet.

Although driver support for Linux is very shallow, compiler support for Windows was abysmal. We did not recompile binaries in Linux to assure fair comparisons between the operating systems, but we were at least given the option. Intel's C++ compiler obviously does not support AMD64 (nor IA-32e) just yet, and no one else seems committed enough to producing a 64-bit compiler for a Beta operating system.

Database Tests Final Thoughts
Comments Locked

29 Comments

View All Comments

  • KristopherKubicki - Monday, July 12, 2004 - link

    tsadowski: Its actually pretty clean. I wouldnt compare it to Gentoo at all. Thats probably also why it has an RPM repository - so you CAN work with it out of the box.

    Kristopher
  • tsadowski - Monday, July 12, 2004 - link

    I find it interesting that you test Fedora Core 2 and expect it all to just work perfectly. Is Fedora not the bleeding edge code, hack it yourself, hobbyist version of RedHat? Not unlike Gentoo? To expect it to just work "out of the box" without some hacking is foolish at least, and at worst perhaps an intentional attempt to slander Fedora/Redhat? I have played around with Fedora Core 1 and while I wouldn't say that it is the best distro I have ever used. I wouldn't bad mouth it without acknowleging it's hackerish, homebrewish nature either!
  • jspaleta - Monday, July 12, 2004 - link

    #15
    along with the flags, the specific versions that you compiled would be good to know. Actually since its compiled I would be interested to know if you had to install/compile any additional build requirements beyond what is available in Fedora Core as well.

    I would also be interesting if you could rerun the
    lame encoding benchmark against the lame build currently available in the stable x86_64 fc2 rpm.livna.org repo, as a comparison.

    -jef
  • KristopherKubicki - Monday, July 12, 2004 - link

    #8, 14: lame and mplayer are compiled. I will get the exact flags and details posted soon.

    #8 again: SuSE gives you two options for installing the drivers - manually, as we did or via YAST. I just chmod 0 /usr/X11R6/bin/X, ctrl-alt-backspace and then run the driver. You can also hit F2 during startup and tell it to go into "failsafe" mode.

    #10: Thanks Matt, id like to work closer with MS to get that. I have a feeling Intel's compiler will show up for x86_64 soon, being as nocona is available now.

    Kristopher
  • jspaleta - Monday, July 12, 2004 - link

    I would like to know from where the reviewers
    got fedora binaries for memcoder, mplayer and lame for fedora core. These untilies do NOT come as part of Fedora Core, are not built by the Fedora Core buildsytem, and can be obtained from a number of different repositories. I would personally like to know if different builds of mplayer/mencoder/lame from different locations experience different results.

    -jef
  • lopri - Tuesday, April 28, 2009 - link

    asdf
    quote:

    quote
  • Term - Monday, July 12, 2004 - link

    Out-of-the-box the video-drivers for Linux from NVidia have Fast Writes disabled, but you have enabled it right?

    damn good article btw
    // Term
  • Possessed Freak - Monday, July 12, 2004 - link

    Errors on graphs:

    Why are the key color orders reversed. Shouldn't red/64bit be on top in the key?

    Why does the order of the OS's change seemingly randomly in the graphs? I thought it might deal with performance, but I could not see a relation.


    Regardless, good article.
  • LostInSpace927 - Monday, July 12, 2004 - link

    I am thinking someone needs to a little research before typing an article.
    "Unfortunately, you can't even try the Personal version of SuSE 9.1 without forking the $90 because the Personal edition does not ship with a x86-64 kernel."
    I downloaded SuSE 9.1 free od charge from www.linuxISO.com.
    All the longer it took me to find this was a few seconds googling.
  • Matthew Daws - Monday, July 12, 2004 - link

    To say that there is no 64-bit compiler for Win64 is slightly untrue: a CPUID.com article uses a beta VC++ 8.0 from the "Microsoft DDK for Windows Server 2003" CD. Sadly, it produces awful code from C++ and cannot optimise less common FPU functions. So, in that sense, there isn't a compiler capable of compiling a whole application: simple benchmarks are possible though (and show 5-35% speed increase, due to more registers mainly).

    Thanks for a good article! --Matt

Log in

Don't have an account? Sign up now