-WinXP vs Win2k3 Server - Benchmarks - DONE

[Nothinman]

if i was an artist, i wouldnt trust an emulated program to run photoshop.

Wine is not an emulator, infact that's what the name stands for.

piracy has nothing to do with me benchmarking it

Sure it does, unless MS gave you a copy to benchmark.

get away from the piracy issue, this isnt about piracy for the 5th time. its about benchmarks and getting the most out of your computer.

It is about piracy, this forum doesn't condone piracy and the fact that you outright said you pirated Win2K3 should get you a temp ban IMO.

And honeslty this thread isn't about getting the most out of your computer, it's about spending time tweaking a Server OS to run like a desktop OS when you could be tweaking the desktop OS. The time you spend making Win2K3 run like XP you could be taking XP and making it run better.

 

[JackBurton]

This topic is so freakin' annoying! Ok, let's say Windows 2003 Server out performs XP slightly in games (which it won't). How would that help someone? Are they now going to go out and BUY W2K3 Server for $1000? YOU DON'T USE A FREAKIN' SERVER OS TO PLAY GAMES! This topic keeps coming up and it is freakin' ridiculous!

 

[NorcoO1]

For F'ing sake, Let him run his benchmarks!!!! You've blown this way out of poportion. I can easily get a Windows Server 2003 120 day trial which is FREE. Why do you assume that if hes benchmarking hes using full blown versions? You guys bitch about him doing something useful with his time, why don't you try the same.

 

[Nothinman]

You guys bitch about him doing something useful with his time, why don't you try the same.

I am, I'm migrating my mail to a new server but I can browse the web at the same time.

 

[batmang]

woot, done. sandra scores significantly better in win2k3 server compared to winxp pro, also runs alot faster/smoother. go win2k3 go.

 

[Sunner]

Well, I guess something good came out of this, you did prove that Win2K3 isn't any better than WinXP for gaming, now all we need is a sticky for this thread so people who wanna run Win2K3 server for gaming can see the benches for themselves.

 

[Twista]

just what i thought since i ran 2k3 before and it does run alittle faster for me but in games it did the same.

 

[batmang]

bump

 

[spyordie007]

Of course you realize that if you "tweak" Windows XP you would likely get comperable results right? I think it only fair that XP gets "tweaked" considering you had to do it to Windows 2003 Server to get it to run the games in the first place.

Other than that the results of this bench are just as I would expect, perhaps this will help to slightly slow down the hundereds of threads we've seen with people trying to run it as a desktop because they seem to think it's "better"...

-Spy

 

[batmang]

i didnt "tweak" anything on any of the os's, like i said above, there from fresh installs with the latest drivers, the only tweaking i did was install refresh lock to rid of the 60mhz refresh rate problem with win2k/winxp/win2k3 server. all i know is, win2k3 fresh install is faster in general then winxp, and thats what i did this benchmark for

*EDIT* - I forgot to mention that i did have to enable Full Hardware Acceleration in the video options to get 3d acceleration to work, its on low by default cause of course win2k3 server is not meant for gaming, its meant for servers, but the refresh lock and enabling full hardware acceleration is all i changed in win2k3 before running the benchmarks.
Fresh installs of the os, latest driver installs, refresh lock then i ran benchmarks, no tweaking was done at all.

 

[spyordie007]

1. 3D Hardware Acceleration
2. Sound

I guess enabling them isnt really much of a "tweak", but than again neither is disabling a service in XP Pro.

-Spy

 

[batmang]

Sound worked fine after installing the drivers from the motherboard cd. But yeah, i know what your saying, i could have just tweaked winxp, and i actually have before and it did improve it a lot. but id rather avoid having to tweak a bunch of stuff just to make it run to its fullest speed. the last time i tweaked xp, i remember it crapping on me whenever id update through windows update. certain things had to be enabled for it to even update, and i didnt even remember the services i turned off, i turned off so many. i used the guide at www.tweaktown.com. their both good os's, but i did get my point across, i think the benchmarks show that.

 

[batmang]

 

[rain2k4]

Whats with all the threadcrapping. Anyone who is on MS actionpack or MSDN could care less how much an individual piece of MS software costs, and if Win2k3 server performs better than XP out of the box, why not use it?

Why would I want to spend ANY time messing with XP if I can get better results using win2k3 out of the box?

Regardless, I dont get why you guys are calling this benchmark meaningless. Now if you guys were as clever as you seem to think you are, why not try and optimize XP and see if you can get the same performance as 2k3? If you can then fine, if not then... stop wasting my time

 

[batmang]

amen.

 

[elbirth]

Originally posted by: rain2k4
Whats with all the threadcrapping. Anyone who is on MS actionpack or MSDN could care less how much an individual piece of MS software costs, and if Win2k3 server performs better than XP out of the box, why not use it?

Why would I want to spend ANY time messing with XP if I can get better results using win2k3 out of the box?

Regardless, I dont get why you guys are calling this benchmark meaningless. Now if you guys were as clever as you seem to think you are, why not try and optimize XP and see if you can get the same performance as 2k3? If you can then fine, if not then... stop wasting my time

Well said. I was reading this thread getting more irritated by the minute... The guy did a harmless benchmark, trying to help others, yet some people can't appreciate that... they have to turn it into a legality issue... they say themselves that any normal user wouldn't spend $1000 on Win2k3 server, so why should he do that just to benchmark it?

In any event, I find this information useful, because at work we may very well upgrade some computers to Win2k3 server for various uses, and I'd like to know what it can do... and possibly install it on my own system if we have the licenses to do so.

 

[metapy]

Originally posted by: rain2k4
Whats with all the threadcrapping. Anyone who is on MS actionpack or MSDN could care less how much an individual piece of MS software costs, and if Win2k3 server performs better than XP out of the box, why not use it?

Why would I want to spend ANY time messing with XP if I can get better results using win2k3 out of the box?

Regardless, I dont get why you guys are calling this benchmark meaningless. Now if you guys were as clever as you seem to think you are, why not try and optimize XP and see if you can get the same performance as 2k3? If you can then fine, if not then... stop wasting my time

Finally the voice of reason, I am tired of people jumping down other's throat for running copy of 2003 Server from their Action Packs or trial copies and "people less kind than most" in this forum completely lambasting them (same people always seem to pour the same rain on all the parades around here). People react as if someone told them there was no god and the mother/wife/sister was an ugly, meretrix all in one sentence when they say they have run, did run, thought about running or heard someone else run 2003 as a desktop OS.

I propose that all the negative comments directed at "TWR2K3ADOS" (Those Who Run 2003 as a Desktop OS) should be morally obligated to provide one UNPUBLISHED tip for XP to make it run better, or some UNPUBLISHED 2003 guide/tutorial/etc. for EACH negative comment made on these boards as they obviously know things we don't.

 

[cleverhandle]

Hate to burst your superiority bubble gentlemen, but if you think these benchmarks show 2k3 to be a better desktop OS, you're an idiot. For the gaming benches - the actual, real-world usage of the OS - the two OS's are already admitted to be functionally identical. That leaves Sandra...

Originally posted by: batmizang
now in sandra, there was a increase in performance in every benchmark i ran.

No. There wasn't. Look at your numbers.

Memory
XP: 5063/5128
2K3: 5121/5080

2K3 takes the first test, XP takes the second. And in both the margin is absolutely insignificant, a whopping 1.1%. Congratulations.

CPU
XP: 9122, 2687/6002
2K3: 9394, 2755/5971

In the Dhrystone - an integer test that has little to do with desktop performance - 2K3 wins by a staggering 3%. Run, don't walk, to the phone and order now. In the Whetstone, which has marginally more to do with gaming performance, the test goes both ways, again by miniscule amounts - 2.5% and .5%. That leaves us with...

File System
XP: 18233kB/s
2K3: 21362kB/s

You at least did a fresh install on the same partition of the same drive, right? Made sure to defragment the filesytem beforehand? Turned off background processes that might be reading the disk? Ran the benchmarks numerous times to report an average and standard deviation or coefficient of variation? Read Eugene's Testbed Article? (Of course you did, it's the one that says MUST READ in big, bold letters.) And you did post these results at Storage Review, right? And were told that Sandra is a completely worthless file system benchmark in practically every way? And then came over to Anandtech because you figured we were all just a bunch of suckers that would assuage your battered ego?

Sorry, try it again with your idiot friends that don't know any better. But it's not going to happen here.

 

[CTho9305]

You need to run each benchmark more than once to get any meaning when the margin is that small.

 

[batmang]

You at least did a fresh install on the same partition of the same drive, right? Made sure to defragment the filesytem beforehand? Turned off background processes that might be reading the disk? Ran the benchmarks numerous times to report an average and standard deviation or coefficient of variation?

And you do all that when you run a game benchmark right? Come on. Like i said, they were fresh installs, and i did defragment, and i did run the benchmarks many times. I didnt touch anything in the services, cause thats part of my point, that win2k3 doesnt have crap running that you dont need by default. Dont you get my benchmark purpose, i stated it at least twice. I ran the benchmarks to prove that win2k3 as a fresh install is faster then winxp pro. And it is. I clearly stated that xp is filled with crap that isnt needed, it slows the system down within windows. yeah sure, you can tweak winxp, but who wants to sit and tweak winxp when a solution could be to just install something else, with the same features, but has the features off that you dont even use. This thread got way more attention then i thought it would, not sure if thats good or bad, but the bashing on me just for running a benchmark is ridiculous. I dont go bashing on peoples 3dmark results asking them for screenshots thinking their lying. You people acusing me are just plain stupid. Your acting like i dont know how to benchmark anything. This is the worst " bitching " ive ever seen from a group of males. Go take some Pamprin, fvcking prissies.

 

[spyordie007]

Chill out, he was pointing out some serious flaws in the data that was given to us from your benchmarking. If you did run the benchmarks multiple times why the single set of data? Nobody is questioning the set of results you posted however he makes a very valid point in that the results you posted did not very signifigantly with the exception of the file system benchmark, however Sandra is not a very good file system benchmarking utility. In all fairness the only really good way for you to do file system benchmarks would be with a seperate newly formatted HD and only after you run tests many times over and average your results since that would give you a better idea of how the OS handles HD thouroput and not your HD.

I ran the benchmarks to prove that win2k3 as a fresh install is faster then winxp pro. And it is. I clearly stated that xp is filled with crap that isnt needed, it slows the system down within windows. yeah sure, you can tweak winxp, but who wants to sit and tweak winxp when a solution could be to just install something else, with the same features, but has the features off that you dont even use.

Windows 2003 Server doesnt even have native support for 3D Hardware acceleration, you have to go in afterwards and install DirectX (this isnt an opional "tweak" this is something you must do if you want hardware acceleration). Or how about turning off that ever-so-annoying-on-a-desktop "why do you want to shut me down" feature?

After you make the changes to Windows 2003 Server to make it act as a reasonable desktop computer you may as well make the changes to Windows XP Pro because what you get after an equal amount of time "tweaking" either of them is essentially the same thing and essentially the same benchmark results (<5% deviation).

You people acusing me are just plain stupid. Your acting like i dont know how to benchmark anything. This is the worst " bitching " ive ever seen from a group of males. Go take some Pamprin, fvcking prissies.

Read the first two words in my post and take a deep breath, the only one here who is attacking anyone else is you.

-Spy

 

[batmang]

*takes deep breath, exhales. ahh.*

i just feel like i cant post something without being... bashed for it. the shutdown thing is annoying, but i just pick something and shut it down. heh. as for directx, with winxp, u have to install a newer directx to play newer games anyway, and most people usually update to the latest anyway, so thats not really something you can compare. enabling full hardware acceleration takes like 5 seconds. so i install win2k3 serv, install direct x 9.0b, and turn on full hardware accelleration. then im all set.

 

[spyordie007]

From the Services MMC:
Disable Fast User Switching
Disable Indexing Service (Which you would have to do for Win 2K3 Server anyways)

From the Desktop:
Change appearance to "Classic"

That will take care of most of it, for the vast majority of people there is no need to go further into it.

-Spy

 

[batmang]

now your refering to winxp services, or win2k3?

 

[spyordie007]

XP, my point was doing those things are just as quick as the things you have done to Windows 2003 Server.

-Spy