Does customization slow down your computer? Answer: Not really

Brad, This is VERY interesting....I know the "MHz myth" was nonsense, but always thought customization scrubed off PC speed/performance - it just goes with the territory. Nice to know that the performance hit is negligible!! If it were, I'd just upgrade to a faster CPU & memory w/FSB. Thanks for this information!!

Well, I guess it depends on your computer. At home on my dinosaure Celeron 333 with only 128 Megs of RAM, I don't run any theme (not even Luna, just plain old Classic), and it does make a difference in speed. But on my work computer (1.4 GHrz w/1G of RAM), it makes no difference in speed whatsoever if I run WB, DX, OB, CX or not. On the other hand WinFX does slow it down.

I run an Athlon XP 2800+, a gig of Geil Golden Dragon RAM and an nVidia FX 5900. Loading *all* the Stardock customization apps not only slows down my computer, but makes it unstable.

I don't think that *anyone* can claim that any PC app does or doesn't do something on all machines. Every setup is so radically different, driver versions, loaded software, brands of hardware, heat, airflow, uptime...all of these things make such a huge difference. Pretending that loading lost of apps on top of stock Windows *can't* slow it down is misleading, and that's what the first sentence of this article implies. The linked page doesn't answer the question for anyone other than Bjorn3D and someone running an exact duplicate of his machine, including all software and drivers.

I think the point is that there is nothing inherent about customization apps that cause the system to perform poorly.

I have little or no performance degradation on my systems from running these apps(WB, WFX, DX, Iconpackager and CursorXP, as well as a few other small utilities), although some of the more complex themes do show a bit more of a hit.

Obviously, drivers and other applications will affect performance. I have also found that it is useful to completely reinstall any version of windows about twice a year (in my experience). Even with restore points, etc., the level of cruft just seems to cause problems over time.

Hey, George Rogers, let's not get into that. Every knowledgable PC user knows that performace does not equal processor speed. It relies on other thing as well, such as cache, bus speed, memory, etc. So no mac bashing please. Although the "MHZ Myth" was a little out of proportion, it was true, and still is .

I'm not sure about that. WindowBlinds, along with Cursor XP and Object Dock and Object Bar eat up about 50MB of my memory. 50MB versus the 0MB extra memory taken up while not running any customization software.

Running all my customising software at once makes no real difference to the speed of my computer, just to the speed of me using my computer, 'cos I keep stopping to look at all the pretty things.

One Internet Explorer process can easily get up past 50MB.

The amount of memory used by any one process (or all of them together) doesn't really matter all that much, unless you're already running out of memory. As long as you aren't constantly swapping out, memory usage doesn't substantially affect overall performance. If you're swapping out, performance really suffers.

CPU usage is really a better indicator of a performance hit.

I don't think there is much of a perf. hit for most modern pc's. However, I have noticed some stablility problems. Mainly w/ object bar. Basically ... major skinning is just not attempted on my main dev machine. On my fun machine however...

Very interesting. I've known for a while now that it mainly depends on your video card. The standard is a 32 Meg 4x AGP. If you are running less than that (like a stock Intel chipset), then OBVIOUSLY your system won't like it. Stock video chipsets are stupid. Meant for nothing more than a 'net box or an office computer, or some computer for your kid. Not for the minimal power user, which is essentially the type of sys you need for stuff like this.

Personally I think this article was crap and only serves to confuse the issue. I mean UT and Nascar benchmarks, give me a break, this isn't another crap Nvidia vs. ATI benchmark, but an attempt to answer the old, "does customization effect performance question" and at that it failed miserably, if one took a moment to objectively look at the tests run and the results of those tests.

Why is it a joke, because sites like Bjorn3D and all other hardware sites know how to do one thing, run video benchmarks for 3D applications, which is what we saw in this little report.

When was the last time anyone saw a decent benchmark of 2D drawing performance? That is what they need to give us, repaint times for windows and your desktop, not just another UT benchmark.

That whole "article" can be summed up as, variance in statistical analysis.

D

The best thing is that in some tests Windows is even faster with all that suff installed..

Windblows definately slows down from customization... But this is what ram is for. I really couldnt say that my computers really slow from little things like themes and icons... I have a 1.0 GHz Athlon with 512 pc133 ram and an oced Radeon 9000 Pro. I know that the biggest hit to performance is spyware and programs that do crap like that...

Stark, your performance hits are probably because of your memory limitations.
If they take up too much memory, then yes, it'll obviously slow your PC down.

To those that think customization slows the system down, where's your evidence?

I've run WindowBlinds, ObjectBar, CursorXP and at different times DesktopX full time while I play games like Counterstrike, Rise of Nations, UT, GTA:VC, etc. and I've never noticed any difference.

This article pretty much just confirms what many of us thought. That customization programs make no more difference than any other process being run on the system. There is nothing inherently slower about customization programs than other other type of program.

It's like those guys who blindly believe that things like msstyles are "faster" because they came bundled with the OS. It's always the quasi-techies who obsess over looking at how much particular processes use in ram while ignoring honking mega hogs like the things that gets stuffed into svchost.exe.

er, frogboy has a point there children...

please forgive the patronisation...but some of us had it coming.

Totally agree Frogboy....my system is only an AMD 1.4Ghz(Palomino)OEM w/512GB of DDR PC2100 which by todays standards is ANCIENT - yet no appreciable performance slowdown!