Preliminary speed test results

I set up the basic installations for my GX150, and got some benchmarks for other distros. Here are some stats to get started:

  • Dell Optiplex GX150, service tag 6DNXL01
  • BIOS version A11
  • Intel 1Ghz Pentium III Coppermine
  • Viking 512Mb PC133 Memory
  • Western Digital 9Gb 7200rpm hard drive
  • 24x TEAC 224E CD-ROM
  • Integrated Intel 815 Series Video
  • Integrated 10/100 3Com Remote Wake-up NIC

As you can see, it’s no speed demon. This is a middle-of-the-road Pentium III with just enough improvements to keep it from being sluggish. In fact, if it wasn’t for the fancy paint job, there wouldn’t be many reasons to keep it around.

I did two installations: the default Xubuntu 6.06.1 and Windows 2000 SP1. I picked Xubuntu over straight Ubuntu because I felt it was a better match for a machine of these specs. And I used W2K because this machine shipped with that version of Windows, and so in a sense, it’s also a good match for the hardware.

Installations were written to a blank, unpartitioned drive from a cold boot. The watch stopped when the machine reached the desktop for the first time and the hard drive came to a stop. (I made that concession to Windows, which plays the trick of showing the desktop before the bootup is complete.)

Since the original poster compared startup times for three programs, I did the same. However, rather than compare IE to Firefox or Abiword to Wordpad, I picked the equivalent for each distro and let the hardware and OS get in the way.

Xubuntu first:

  • Default Xubuntu 6.06.1 installation
  • Installation time: 00:28:30
  • File system: default ext3 partitions
  • Cold boot startup time: 00:01:04
  • Shutdown time: 0:01:29
  • Abiword 2.4.4 startup time: 0:00:04.0
  • Gimp 2.2.11 startup time: 0:00:13.1
  • Firefox startup time: 0:00:08.8 seconds
  • Net speed test: 1.7 megabits per second (average of 3 tests)

Start times and shutdown times are abysmal, that much is obvious. And the startup time for Firefox is poor by any measure. It shouldn’t take that long for a browser to start up. Or at least that’s the consensus.

One other note: Just to level the playing field, I disabled the login window and booted straight to the desktop. And I felt it necessary to change the color depth since I get screen artifacts at the default depth. That, however, I blame on this machine and not on Ubuntu.

Now Windows:

  • Default Windows 2000 SP1 installation
  • Installation time: 00:34:00
  • File system: default NTFS partition
  • Cold boot startup time: 00:01:03
  • Shutdown time: 00:00:21
  • Abiword 2.4.1 startup time: 00:00:05.1
  • Gimp 2.2.9 startup time: 00:00:11.1
  • Firefox startup time: 00:00:09.5
  • Net speed test: 1.1 megabits per second (average of 3 tests)

Start time, interestingly enough, is almost identical. And while the Gimp started faster (yes, I did the initial setup before testing the speed), both Abiword and Firefox opened faster on Ubuntu. That’s an interesting note, considering Ubuntu’s reputation for sluggish browser setup.

The installation time for Windows also includes the time it takes to format the drive in NTFS. Even though the drive was blanked beforehand, W2K SP1 doesn’t offer a “quick format” option like XP does. A lot of people forget that. You might be able to rationalize that W2K’s installation time is longer for that reason, and I do too. But to me, logic says If it wasn’t the only way of doing things, it would be faster.

Abiword, the Gimp and Firefox all had to be installed as third-party applications. (No, I didn’t time IE. There’s no IE in Ubuntu short of Wine, so there’s no comparison to be made.) Abiword and the Gimp were installed off the OpenCD v3.1, while Firefox was downloaded.

I can’t account for the drastic measure in net speed. My only guess is the traffic load at different parts of the day.

One last thing:

On the surface, those things look rather interesting. But as I’ve mentioned in the past, I think Windows users overlook one last point: They’re not done yet.

Installation took 34 minutes. But this machine, with a default W2K install, I had

  1. no Internet access until I had gone through the configuration wizard;
  2. no video drivers, and therefore
    1. no native resolution, and
    2. an unusable color depth;
  3. and no sound.

So the three main things I might use a computer for (Internet, video and sound) are all missing. And by missing I don’t mean you have to download a codec to get your mp3s playing (like Ubuntu), I mean there’s no sound.

So yes, installation only took six minutes longer. But you still have to configure your connection, download video drivers, download DirectX and download audio drivers before you’ll even hear the startup tune play.

(I’ve seen worse, too. I have a GX260 that uses onboard Intel 915 graphics. A clean XP install gives me 480×320 at 4-bit color depth and no network connection whatsoever. And that’s a legitimate XP installation, too.)

And of course, where Ubuntu is (for all practical purposes) impervious to standard-issue viruses and malware, W2K still needs spyware scanners, antivirus programs, tuneup programs and whatnot. That’s when the numbers really start to go up.

Now for the disclaimers.

Every machine is different, every part is different, every installation is different … and even every environment is different. You could make the case that your computer might start up faster in winter than in summer, and I might buy it.

But a couple of things stand out to me. First, yes, I concede: Ubuntu shuts down slower. However, in my experience, Windows 2000 starts up just as slowly. The Gimp starts up faster. But Abiword and Firefox start up slower in Windows.

And when you install Ubuntu, you’re done. You install Windows and you’re only halfway there.

Next stop: A custom Xubuntu installation with all the bells and whistles I can think of.


One thought on “Preliminary speed test results

  1. Pingback: I think they’re all gone « Motho ke motho ka botho

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s