Vista Ultimate vs Server 2008 Workstation

TwistedJester

Weaksauce
Joined
Jun 15, 2005
Messages
73
I'm looking to make the OS jump since I'd like to bump myself up to 4 gigs of RAM for gaming, but I have two options available to me. On one hand, I can get Vista Ultimate from my university for free. On the other, I can get Server 2008 from Dreamspark for free. Looking around, it seems that in most benchmarks, 2008 outperforms Vista Ultimate despite having the same kernel. I know I could get both and mess around with them, but I'd really just like to get one and be done with it. So can anyone throw in their two cents regarding which version I should spring for?
 
If it is for a workstation/non-server PC, stick with Vista Ultimate. The performance difference, if any, isn't enough to justify the hassles of extra tweaking an configuration to get basic things running on Server 2008 that are already done on Vista. They also aren't enough to justify the hassles of finding enterprise level software to handle system maintainence. If the job calls for a screwdriver, don't use a hammer.
 
I found Server 2008 definitely snappier (installed on same exact laptop that Vista was). And didn't have to do any tweaking or whatever to get it to do what I needed....it's like Vista already running in Performance Mode.(so actually less steps versus Vista)

However...since you mention gaming, I'd stay away from Server OS. You may run into an install error..some apps won't install on a server OS. You may also find some quirk about drivers, it's sometimes best to stick with drivers rated for a server OS (instead of shoehorning a Vista video driver into it), that won't give your game what it needs. The server is tuned for background services 'n such, not foreground apps.
 
Installing a server based OS for a non server function is, in my opinion, asshattery to the extreme. And they say Apple zealots are dumb. 'Look at me! I Are RunNinG teh SURVUR OS cuz I iz leets'! snore.....
 
Installing a server based OS for a non server function is, in my opinion, asshattery to the extreme. And they say Apple zealots are dumb. 'Look at me! I Are RunNinG teh SURVUR OS cuz I iz leets'! snore.....

Don't knock it til you try it, it's fast, and it's stable. It's been quite common with IT people for the past few versions of Windows too. More importantly you get more built in useful tools and utilities over a desktop OS. With IT people, spare licenses of Microsoft products are as abundant in the office as USB cables, spare mice and patch cords....so it's nothing about using a 7 year olds terms like "look at me I'm l33t". :rolleyes:
 
I haven't tried to use 2008 as a workstation yet, but I have used NT4 server, 2000 server, and 2003 as a workstation in the past. It has a tweaked kernel that utilized the ram better. However, I alway thought Vista finally utilized the ram cache as the maximum amount until the application is loaded.
 
I found Server 2008 definitely snappier (installed on same exact laptop that Vista was). And didn't have to do any tweaking or whatever to get it to do what I needed....it's like Vista already running in Performance Mode.(so actually less steps versus Vista)

However...since you mention gaming, I'd stay away from Server OS. You may run into an install error..some apps won't install on a server OS. You may also find some quirk about drivers, it's sometimes best to stick with drivers rated for a server OS (instead of shoehorning a Vista video driver into it), that won't give your game what it needs. The server is tuned for background services 'n such, not foreground apps.

Usually you have a few things you want to enable with server os to use it as a workstation but it doesn't take much.

The big issue is app support. A lot of programs will not install on a server os. Antivirus, disk defragment software, etc usually has a special server version.

My sprint evdo card even cares. The old version of the software works fine in 2003. The newer version checks and will not install.
 
The big issue is app support. A lot of programs will not install on a server os. Antivirus, disk defragment software, etc usually has a special server version. .

I agree with you. But then us IT propeller heads usually have the licenses and resources for that. Good ol NOD32 2.7 went right on mine, etc.

In the OPs scenario though, for gaming, I'd discourage Server.
 
I agree with you. But then us IT propeller heads usually have the licenses and resources for that. Good ol NOD32 2.7 went right on mine, etc.

In the OPs scenario though, for gaming, I'd discourage Server.

Yea 2.7 didn't care but 3 does(I know you don't like 3). Most others get pissed unless you are running the business version.

Can't speak for 08 but 03 has always had issues with the power saving mode on my notebook as well. I don't let it go to sleep so I've never cared.
 
always had issues with the power saving mode on my notebook as well. I don't let it go to sleep so I've never cared.

...same here.

Yeah, not crazy about 3.0...only have it on 2 of my rigs (both happen to be Vista, third Vista rig still on 2.7). And only have it deployed at 1 law firm client..about 1/2 of their rigs. Not installed on any servers....still too many issues w/it.
 
Here, here and here. There are more if you make the right google query.

This is interesting. With my test laptop, using Vista Business x64 SP1 and Server 2008 x64 Standard, using the *exact* same drivers on each, my 3DMark06 scores were within single digits of each other. I would have tested 3DMark01, except it didn't seem to run under Server 2008 for some reason.

I also ran Passmark, which showed a tiny score difference (mostly related to not running Aero in 2K8) and did some network file copying tests, which did not show any meaningful difference. The two operating systems performed the same among all tests I tried.
 
The two operating systems performed the same among all tests I tried.
That was my experience as well, using a Core2Duo tower and a removable SATA drive cage. Even the two hard drives I used for the test were identical models and firmware revisions.
 
Why chose? You can get both for free, then for Gods sake, build two computers! :p
 
Usually you have a few things you want to enable with server os to use it as a workstation but it doesn't take much.

The big issue is app support. A lot of programs will not install on a server os. Antivirus, disk defragment software, etc usually has a special server version.

My sprint evdo card even cares. The old version of the software works fine in 2003. The newer version checks and will not install.

I have both at Home. I agree with this post. If you are using it as a Workstation/Playstation you should use Ultimate. You'll have issues with Server 2008 as mentioned in this Post. I wouldn't want to try and Game on it. If your Idea of a "Game" is Terminal Srvcs, and DFS, AD Server, etc, then install Server 2008.
 
I'm looking to make the OS jump since I'd like to bump myself up to 4 gigs of RAM for gaming, but I have two options available to me. On one hand, I can get Vista Ultimate from my university for free. On the other, I can get Server 2008 from Dreamspark for free. Looking around, it seems that in most benchmarks, 2008 outperforms Vista Ultimate despite having the same kernel. I know I could get both and mess around with them, but I'd really just like to get one and be done with it. So can anyone throw in their two cents regarding which version I should spring for?


If you are not going to run 6 or more gigs, you are most likely wasting your time with 64 bit. The 64 bit instruction set makes your RAM reqs go up. Most systems I've seem start to hit their stride at 8 gigs or more. I think I've seen RAM requirements go up 60% with 64 bit instruction sets. So I wouldn't be surprised if a 4 gig 32 bit system, with all the issues that entails, and a 4 gig 64 bit system, that the 32 bit system would perform better. Comments.
 
So I wouldn't be surprised if a 4 gig 32 bit system, with all the issues that entails, and a 4 gig 64 bit system, that the 32 bit system would perform better. Comments.
I would be quite surprised if the x86 performed better. A x64 OS will have slightly higher memory usages than it's x86 equivalent, but the fact that you can use all 4 GB rather than 3.4 or so easily makes up the difference. If you simply install Vista x86 and Vista x64 on the same hardware, you'll see a minor difference in speed, and a very minor difference in memory usage.
If you are not going to run 6 or more gigs, you are most likely wasting your time with 64 bit.
I'm running 4 GB in three different systems, and I'm hardly wasting my time running Vista x64 on all of them. It sure as hell beats wasting some of my memory.
 
I recently built a small server for home (p5q pro, 8gb ram, raid0, 8800gt) and have been running server08 on it. Everything to me feels a bit faster. With that being said, I installed the new brothers in arms last night and gave it a go. The game was very skippy/hiccupy. I then installed vista64 and it played perfect. I didn't have much time to look into this, but I'm sure theres something I can do to tweak 08 into not skipping so much. nevertheless, there's a real-world review :)
 
I would be quite surprised if the x86 performed better. A x64 OS will have slightly higher memory usages than it's x86 equivalent, but the fact that you can use all 4 GB rather than 3.4 or so easily makes up the difference. If you simply install Vista x86 and Vista x64 on the same hardware, you'll see a minor difference in speed, and a very minor difference in memory usage.

I'm running 4 GB in three different systems, and I'm hardly wasting my time running Vista x64 on all of them. It sure as hell beats wasting some of my memory.

The info is out there, I'm not making it up, read Tom's Hardware reviews on Vista 32 and 64 in reference to memory . Sorry to ruin your day or pop your bubble. The difference in RAM is not minor unless you call a 60% hit minor.
 
I recently built a small server for home (p5q pro, 8gb ram, raid0, 8800gt) and have been running server08 on it. Everything to me feels a bit faster. With that being said, I installed the new brothers in arms last night and gave it a go. The game was very skippy/hiccupy. I then installed vista64 and it played perfect. I didn't have much time to look into this, but I'm sure theres something I can do to tweak 08 into not skipping so much. nevertheless, there's a real-world review :)

Server feels faster because it has been stripped of a lot of the candy. If you don't load up a ton of services it will outperform Ultimate.
 
Tom's Hardware
:rolleyes:
The difference in RAM is not minor unless you call a 60% hit minor.
If it truly was 60%, then we'd have something to talk about. Fortunately, it isn't so in actual usage. On my Dell Latitude D630, with 2 GB of memory, I can barely see a difference in memory usage between Vista Business x86 and x64, and that's with far less than 4 GB of memory. Try it out, instead of buying into the hype that a questionable website is telling you.

If not, just think about it reasonably. Would so many of us be using Vista x64 if it used 60% more memory? I don't think so. Common sense.
 
:rolleyes:

If it truly was 60%, then we'd have something to talk about. Fortunately, it isn't so in actual usage. On my Dell Latitude D630, with 2 GB of memory, I can barely see a difference in memory usage between Vista Business x86 and x64, and that's with far less than 4 GB of memory. Try it out, instead of buying into the hype that a questionable website is telling you.

If not, just think about it reasonably. Would so many of us be using Vista x64 if it used 60% more memory? I don't think so. Common sense.


http://www.tomshardware.com/reviews/vista-workshop,1775.html


I backed up what I posted with a source.

I have a 4gig Vista Ult 64bit system, that I can't afford to throw out the memory and buy 8 GB of DDR3 1800 mHz CAS 7 or I would.

Oh, and I stand corrected, it is 20-40% more memory depending on the application.
 
Server feels faster because it has been stripped of a lot of the candy. If you don't load up a ton of services it will outperform Ultimate.

That is correct. After loading a few services it does start to bog it down a bit.
 
Oh, and I stand corrected, it is 20-40% more memory depending on the application.
So you chastise me, and then correct your previous statement? That's freakin' hilarious! You backed up what you posted with a link that says something different....nice.

Somewhere in the realm of 10 to 20% could be believable....60% isn't even logical. Just because Tom's Hardware posts a link about something, doesn't make it so. I can find you a link that says aliens are running the White house, and that doesn't make it so (although it would explain a lot of the past 8 years). Anandtech, which is a far more trusted website than Tom's Hardware, used to have links showing conclusively that RAID0 wasn't really all that special on a desktop, and yet that debate raged on for years.

That's why I suggested trying it out yourself. Nothing trumps first hand knowledge, so find yourself a system capable of running both platforms, and give it a whirl.
 
I hate to get in the middle of this but this article from Tom's is BS, a whole page dedicated to disabling hibernation? WTF? More and more Tom's is reminding me of magazines like Sound and Vision, "I can just feel the system is faster", then where are all the benchmarks backing it up? One page of 3 graphs is all you can muster while you can go on and on about memory address mapping and cheap ram. This is why I've always read [H] since way back when, because they go through everything thoroughly; no crap about how it feels or how it sounds, I want charts and graphs and hard numbers backing up what's said.
 

If you would close your mouth long enough to read what Deacon said, his knowledge is firsthand experience.. you know, like, actually using the fucking computer.

How does one link that?

You on the other hand spew third party crap and have no hands on experience. You even proved you didn't know what you were talking about by spitting out numbers randomly.

This is a hardware site in case you didn't notice. A lot of us have real experience with computers.

Knock it off with the age comment already. It's getting old.

And by the way, I'm going to back Deacon up. When I first got Vista from the Microsoft Feedback Program, I installed the 32-bit version and enjoyed it for a couple months. Then I decided to take the 64-bit plunge. There was NO real-world difference in speed between 32-bit and 64-bit. NONE. I don't need Tom's line-my-pocket-with-money-and-I'll-lie-for-you Hardware to tell me I'm wrong. I know I'm not.
 
Tom's is BS
That's why I had to laugh from the very beginning, when I saw Tom's Hardware was the source. For the right price, Tom will tell you that a Pentium 3 is faster than a Core 2 Duo, because 3 is higher than 2.

That's why I made the suggestion to try it out yourself. You will use a little more memory on the x64 side, for several reason, one including the software itself. If you look at drivers and applications, in terms of file size, the x64 equivalents tend to be a little larger. I wasn't disagreeing that x64 does take up a bit more memory, I was taking issue with the 60% comments....which are a moot point now anyway.

Furthermore, if you take a system with 4 GB of memory, I'd gladly use Vista x64 on it instead of Vista x86. That slightly higher memory usage would disappear considering the fact that you'd be using all 4 GB, rather than 3.4 GB, or less in many situations.

So, in terms of being constructive, I think it is safe to revise the comments given much earlier. If a system has 2 GB of memory, you aren't really gaining anything by running Vista x64, not the 6 GB as suggested before.
 
If you would close your mouth long enough to read what Deacon said, his knowledge is firsthand experience.. you know, like, actually using the fucking computer.

How does one link that?

You on the other hand spew third party crap and have no hands on experience. You even proved you didn't know what you were talking about by spitting out numbers randomly.

This is a hardware site in case you didn't notice. A lot of us have real experience with computers.

Knock it off with the age comment already. It's getting old.

And by the way, I'm going to back Deacon up. When I first got Vista from the Microsoft Feedback Program, I installed the 32-bit version and enjoyed it for a couple months. Then I decided to take the 64-bit plunge. There was NO real-world difference in speed between 32-bit and 64-bit. NONE. I don't need Tom's line-my-pocket-with-money-and-I'll-lie-for-you Hardware to tell me I'm wrong. I know I'm not.

I don't get your post at all. He says there is a big difference in his post, you say there is no difference.

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Quote:
Originally Posted by les_garten View Post
So I wouldn't be surprised if a 4 gig 32 bit system, with all the issues that entails, and a 4 gig 64 bit system, that the 32 bit system would perform better. Comments.
I would be quite surprised if the x86 performed better. A x64 OS will have slightly higher memory usages than it's x86 equivalent, but the fact that you can use all 4 GB rather than 3.4 or so easily makes up the difference. If you simply install Vista x86 and Vista x64 on the same hardware, you'll see a minor difference in speed, and a very minor difference in memory usage.
Quote:
Originally Posted by les_garten View Post
If you are not going to run 6 or more gigs, you are most likely wasting your time with 64 bit.
I'm running 4 GB in three different systems, and I'm hardly wasting my time running Vista x64 on all of them. It sure as hell beats wasting some of my memory.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

Now you boys can keep giving each other a reach around if you want.

I can prove I've deployed over a billion dollars worth of Hardware and software over the last 23 years(as director of Data Engineering Teams, as well as CTO for other endeavors). So you really don't know anything about my experience.

One thing I did do is say here are links to back up my info.

You two nut jobs just said, I say it's so.

I made a post, it was argued with so I went out and posted links to back up my post. You guys with your panties now fully wadded up, just say, well we say it's this way. And neither one of your posts agree with each other! Boyz, leave those Weasels alone!

Everything I stated is well documented on the web. It's a simple statement, don't go way out of your way with big expectations if you have 4 GB of RAM. Because of the instruction set memory addressing overhead of 64 bit, that that 1/2 gig is going to get eaten up by the 64 bit OS and the app being run. If you chunk in some big RAM (8 GB), 64 bit makes sense. If you've used em all, you know the software and hardware issues with 64 bit. At 4 GB, it ain't worth the transition, this comes from my EXPERIENCE. I just happened to back it up with references.

Some of my references show that a lot of apps run faster on 32 Bit Vista because you don't have the WoW overhead.

Now if you two rubes got anything to contradict that statement, you'll need something besides, "me and my banjo pickin' brother say so".
 
les I still don't get your deal, and reading along I'm trying to figure out are you saying that x64 is only for those with above 4gb or are you saying that x64 will slow you down if you only have 4gb or is it both? ZDNet says here (http://blogs.zdnet.com/hardware/?p=2354) that even 8gb of ram is kind of moot unless it's for VM's and Photoshop, even then anything A/V with a scratch disk should be pushed to a RAM disk but that's a whole other can of worms. Also the links you posted, the vistax64 links to Tom's, the ZDNet said "Measuring memory use is tricky. For example, the x64 systems occasionally grabbed extra RAM from the available pool, presumably to use in building the page file and Superfetch cache. In all cases, though, memory usage eventually retreated to a baseline level and stayed there." ExtremeTech said it's a toss up between all the apps and 4sysops is all over the place but says you need over 4gb for VM's and testing and under 3gb with all 64 bit apps but they also are second hand info from C't which I would like to see the original article to more understand them. I'm just trying to figure out what your stance on this is.
 
If you are not going to run 6 or more gigs, you are most likely wasting your time with 64 bit.
Do you like using a product you bought? Sorry, but I like using the memory I purchased, which makes 64 bit necessary.

Server feels faster because it has been stripped of a lot of the candy.
Funny, considering the fact that Ultimate offloads all the Aero to the GPU and has no effect on your processor whatsoever...


A CTO that relies upon Tom's Hardware as his factual evidence?
L.M.A.O.

You, sir, have no idea what you are talking about.
 
Hi Mr White,
My stance addresses an amalgam of all the issues. I'll try to draw this out as clear as possible. Here are the issues I was addressing.

1) Someone says, "I have 4 GB, I must use 64 bit Vista" , this is VERY prevalent misconception everywhere.
2) Vista 64 is Faster than Vista 32
3) Vista 64 is slower than Vista 32

So I'm saying if you have 4 GB and you(a customer or a friend) said you were leaving performance on the table and had to go to Vista 64(because Vista only showed 3.4 GB), I would point out these articles that show that with some appz, V-64 runs slower than V-32.

That the Memory Remapping in the Bios may not be such a huge performance hit. That V-64 uses 20-40% more memory(depending on the app), so the 600mb difference in RAM is not that huge of an advantage considering WoW and the rest of the 64 bit Vista headaches, drivers, driver signing, software availability, etc.

I would also say that if you think you are going to do the Vista 64 deal, give it what it needs to let it run. As per what I said, and the article you posted, it needs more than 4 gb to really do anything over 32 bit or even itself. So that's where I came up with 6 GB, and ideally 8GB to match all four slots.

IF I had DDR2 in my box, I would throw 16 GB to it or at least 8 GB. But I've got 4GB DDR3 very fast Cell Shock 8-7-6-21 1800 mHz and I just can't justify pulling it all out and dropping another $700 bucks on memory just to play with it right now.

In my home Lab, I am running V-32 on my MacBook Pro, V-64 on My Desktop with the DDR3, and 2008 on another Older P-4 Prescott box. As well as Solaris, IRIX, BSD, and all the other Fun Stuff just to see the issues with them all because like a lot of folks here, that is my business.

I think my advice for most bang for the buck and ease of use is 3-4GB in a V-32 box. I would put in 4 GB just to balance the slots and pickup the other 4-500 MB over 3GB. Then just be slightly irritated at Windows only Showing 3.4GB. The only other wildcard here would be a Big RAM Video Card that could grab a lot of Remapped RAM like the one I have 2 GB 4870x2.

For a tinkering Power User, Skulltrail, Twin CPU, 32 GB, V-64!

Hope that makes it clear, my Snarky comments were tongue in cheek and meant to be a little on the fun side. I really don't feel we're that far off, except where I originally came from which is V-64 ain't nothing special at 4GB compared to V-32. If you're gonna eat the Headaches of v-64, Grab some RAM and do it right so you MIGHT at least see a little performance. This was really meant to address all the Noobs who say I got 4 GB and HAVE to do V-64. The issue is a lot more complicated than that. That's it.
 
Keep the conversation civil please.
Namecalling and insults will only result in infractions.

Thank you.
 
Thanks les for putting that all into perspective, it was getting confusing reading through the thread of comments.
 
I obviously read a lot of things, like for instance here.

The ongoing theme here, I presented my view and evidence to support it, and basically received nothing but ridicule, no evidence. Like I said, you have controvertible evidence, let's see it.

Admittedly, Tom's is not what they were back in the day. That doesn't mean that article I sited doesn't make sense, especially when backed up with a number of other sites. Come on let's see something?
 
I have a box running an Intel E2180 processor, 4GB of RAM and an nvidia 8800GTX. right now it's dual booting Vista 64 Ultimate and Windows 2008 Server.

I've tweaked out 2008 using the various online guides to make it a decent desktop OS. It seriously takes a bit of tweaking and even downloading of components to get it just right. When all is said and done, on this particular PC, I am not seeing any performance differences between either OS on the same machine.

If your goal is to learn 2008 server and use some of its features, and there is a lot to like about Win 2008 server, then you can definitely benefit in using it over Vista.

If your goal is to get more performance, and to simply avoid Vista, you're really not going to see any real differences other than 2008 being a bit less user friendly than Vista.

If you have a specific NEED for 2008 server, then use it. If not you will be far better off using Vista due to fewer compatibility issues, ease of use, and included features.
 
Back
Top