speed dif between xp and win98?

[N.V.M.]

has there been any benchmarking of the speed of the client using different o/s's? this is kind of critical for me to know this because if , lets say, xp was noticably faster, i'd have to invest in larger hard drives to handle that size program.

[iggy]

The client can't be bencmarked in Win9X OSs. There have been posts about client performance (search might help) in some other (mainly Linux) OSs, in comparison with WinXP - only Gentoo distro managed to beat WinXP.

[FoBoT]

something to keep in mind is the amount of ram

if you build two identical 128MB RAM boxen , put XP on one, Win98 on the other , and use the -rt switch on both, the XP box will start running out of memory and paging sooner due to the extra RAM requirements of XP

now if these alleged boxen have more than 128MB of RAM , this doesn't really come into play at all

i have some 128MB lab boxen that aren't used part of the year, when not in use for "real" work running XP, i often put win95/98 or linux on them to improve production , i don't have any hard/fast statitstics/benchmark info that i can give you though

[Paratima]

Yup, there IS a lot that goes into it and unless you were comparing different OSes on the SAME box, in a no-net environment, I don't think I'd trust the numbers. For instance, Windoze tends to "chat" on a network a lot more than Linux. Just "Hi, how are ya?" kinds of messages. Enough to make a difference? Who knows? How heavily populated is your LAN?

Sometimes the buffered hard disk write on NT/W2K/WXP can help. Sometimes it hurts!

FoBoT mentioned RAM. Hard drive characteristics come into it too. FAT32 or NTFS or one of the Linux variants? How well de-fragged?

It all just boggles my poor tired brain. One piece of advice, though. If you DO get it figured out, you'll be a lot better off to keep it to yourself. There are TONS of uninformed, semi-informed, and well-informed opinions floating around out there. Anything you say in a forum like this will start a major debate.

[iggy]

Agree completely with FoBoT and Paratima - own experience is necessary.

Anything you say in a forum like this will start a major debate.

Isn't debate what the forums are about?

[N.V.M.]

so nobody knows,hmmmm. if i had a system with lots of ram(re:memory not a factor), there should be an answer.

my point was to try to save the cost of buying larger drives for win xp, but i'm not willing to do that and find there's no real difference.

i'll keep researching though.

[tpdooley]

For phase I, a default Win98 setup was somewhere between 5-10% faster than a default WinXP install. By following directions to eliminate a number of unneccesary apps running in the background, it was claimed you could get WinXP to run at the Win98 speed.

There were claims that Linux was faster than windows - yet several that tried it out mentioned not being able to see a performance increase. If you know what you're doing, eliminating the unneccesary apps in Linux will also improve the performance. (Find the fanatics on your team that have lots of experience with that OS..

I'd suggest that instead of wasting money buying larger hard drives and winXP upgrades - you'll be better off buying additional new systems that will increase your production rate. i.e. buy a 5th system, and increase your production by 25%.

[N.V.M.]

ummm,you mean an 8th system.....lol


thanks for the response tp, good stuff!

[rstarr]

I've been working with DC projects for the last 2 1/2 years using anywhere from 14 to 50 pc's which have basically the same hardware.

I've used Win95, Win98SE, Win2K, WinXP and Linus flavors. It's really hard to draw conclusions as to "what is better"

It really boils down, my opinion, to minimum hardware configuration. Having at least enough ram 256 - 512 meg, a hard drive large enough to hold your os, proggies with at least 25-50% free space, good quality hardware (motherboard, cpu, PS and video card)

This is what I found that makes for a nice average system:

30-40 gig 7200 rpm hard drive, 512 meg of DDR ram: PC2700 or higher, AMD XP2100+, WinXP OS (stripping out the un-needed background proggies), video card, at least AGP4X DDR based 64meg.

Go ahead and fire away guys and gals.

[Ned]

Back in Phase One, I had a machine with Win'98, W2K, and Linux all installed.

Linux produced 13% more than Win'98 which in turn was about 10% better than W2K.

So, same hardware with these operating systems produced the above results. Same code in the case of Win'98 versus W2K.


Ned

[FoBoT]

i have a lot of similar boxes here at work, if i get some free time in say, december or january, i can install 95-98-w2k-xp and any linux distros on identical boxen and benchmark them

somebody remind me if i forget, too much "real" work going on right now

[N.V.M.]

there is no doubt,after trial and error, that the difference between win xp and win 98 is LARGE. i haven't tried w2k, but man, the work i wasted using win 98 all this time.
i was forced to use win98 to begin with on my farm(small hard drives) but after picking up some bigger storage and installing win xp,

[jonnyw]

Originally posted by rstarr
...WinXP OS (stripping out the un-needed background proggies)...

fancy sharing with us which ones are unnecessary (Preferably with a small description of what each does, so we can be sure its unnecessary for us as well)?

[IronBits]

Go here http://www.blkviper.com/WinXP/servicecfg.htm
or here http://www.tweakhound.com/xp/xptweaks/supertweaks1.htm

[jonnyw]

cheers ironbits, will look into those tonight