XP vs Vista vs 7: Resources + More

Kyle

New Member
#1
Okay, so there's been a lot of talk about which is better, which runs leaner, which takes more space, etc.
So I decided; Hey! Why not test them all, in a controlled environment, allowing them the exact same treatment to get as accurate a result as possible? Good Idea?

I think so too. So here comes Kyle's Windows XP/Vista/7 Resource Test + More!


Environment

First things first, the environment I used.
Core 2 Quad Q6600 @ stock (2.4 GHz, I reset the clock to get an accurate test)
3 GB System RAM (800 MHz DDR2)
ATI Radeon HD 3450 256 MB @ Stock (Again, reset)
Asus Motherboard
Integrated 10/100/1000 LAN connection
Windows 7 Build 7000 Ultimate x64 as the base OS


Then, I installed Microsoft Virtual PC 2007 (64 Bit installer) on that platform, with nothing running in the background.
I then allocated 1 GB of RAM for the Virtual PC
I then set up three disks, One 10 GB disk for XP, and two 20 GB disks for Vista and 7.
The virtual machine itself only uses one core, so I let it have the full 2.4 GHz core.
I then installed each OS (Using XP Pro SP3, Vista Ultimate SP1, and The Windows 7 Ultimate Beta, all 32 Bit OS')
Then, I booted into each separately and ran the First boot memory tests.



Memory Test Results:

Windows XP Professional SP3 Results:

No Optimization (Untouched): http://img22.imageshack.us/img22/6011/xpsp31ncfp1.jpg

Optimized (All unnecessary processes and services disabled): http://img17.imageshack.us/img17/7503/xpsp32cfk5.jpg


Windows Vista Ultimate SP1 Results:

No Optimization (Untouched): http://img18.imageshack.us/img18/6092/vistasp11ncyx8.jpg

Optimized (All unnecessary processes and services disabled): http://img24.imageshack.us/img24/5572/vistasp12ckx0.jpg


Windows 7 Beta (Build 7000) Results:

No Optimization (Untouched): http://img24.imageshack.us/img24/6951/7beta1ncwl2.jpg

Optimized (All unnecessary processes and services disabled): http://img6.imageshack.us/img6/6461/7beta2cxi1.jpg


---------------------------

Conclusion:

That's it so far, showing that (obviously) XP SP3 has the lead in RAM usage (76-82 MB), followed by 7 Beta (288-320 MB) and Vista SP1 (295-347 MB). Overall XP also has the least bloat, with 6.3%, followed by Seven with 10% and Vista bringing up the rear with 15%.

---------------------------


Hard Disk Space Test:

Windows XP Professional SP3 Results: http://img10.imageshack.us/img10/4320/xpsp3hdfitestmy5.jpg

Windows Vista Ultimate SP1 Results: http://img24.imageshack.us/img24/1181/vistasp1hdfitestgm4.jpg

Windows 7 Ultimate Beta Results: http://img18.imageshack.us/img18/3/7betahdfitestbk0.jpg


---------------------------

Conclusion:

Again, XP comes out on top, its clean install footprint only being 3.68 GB. Second place is 7 Beta, with 6.09 GB, and Vista comes in last with 7.34 GB. Seems like Vista has the most bloat on the hard disk as well. As for direct comparison, if XP = 100%, 7 Beta comes in at just over 65% more bloat, and Vista brings up the rear with over 99% over-sized.

--------------------------

More tests coming soon, suggestions?
They can't be graphic based, due to the limitations of a virtual machine, but any other test I can administer easily.

.
 


loathe

New Member
#2
You have made a very good point the Windows 7 Beta is the first OS to use less system resources the it predecessor.
Lets hope it stays that way in the final release
 


davehc

Essential Member
Premium Supporter
#3
Your tests are really a little meaningless. The whole point of Vista and, improved, 7 was the availability of resources. This meant loading more DLL's etc at the initial start.
As we move on to updated OS's, and this will include Linux distributions, it is obvious that more demands will be made on hardware.
The memory allocations, in the tests, do not appear to be paramount even with 3Gbs. Xp is, of course, very low, but has poor memory management and will load up to a higher figure when using apps.; Vista was similarly inclined, but handled the memory a litle better. Windows 7, even in Beta, is exrtemely good at handling it. Solution, as we move along to new OS's (Its going to be even more demanding if anyone can afford a touchscreen) Put more ram in.
Hard disk space? Are you serious? When Xp was being planned, most hard disks were really small. Today, the difference in the "footprint" is not even a consideration. One point there. Many users in the XP days, slowly had to update their hard disks, as , not the OS, but most modern software began to bloat. One example springs to mind (there are many) Nero.
Fwiw. I just checked my Vista installation against Windows 7. Running the same programs, and with this site open, Vista is using 284mbs and Windows 7 is using 295mbs.

But, maybe we should all look at this. Remember this was an early release, with compatibilty and other issues. Win 7 beta has improved quite a lot since then.
http://blogs.zdnet.com/hardware/?p=3187
 


Kyle

New Member
#4
Fwiw. I just checked my Vista installation against Windows 7. Running the same programs, and with this site open, Vista is using 284mbs and Windows 7 is using 295mbs.
Are those both first boot results, or did you let them sit?
If so, superfetch may skew your results.

As for the relevancy of my tests, you're right, for any user with a computer bought in the last two years this probably won't matter too much, but if you've got an aging computer with little RAM and a small HDD, this could mean the difference between sticking with XP and moving to 7.

But it still may be a consideration when users go to update, regardless of specs. Most people end up adding things like Antiviral software, firewalls, MSN, Yahoo Messenger, Quciktime and iTunes helpers, Torrent software and more to the boot up, and when you've got all that junk running, the quicker and less resource hogging the OS is the better that computer's going to perform with the tasks of loading and running those programs.
 


Kyle

New Member
#6
Lmfao.
IMO, you can't really compare Windows to Linux or Mac OS, at least in speed.
They're very differently put together. It's like organic apples and genetically engineered giant robot apples. Haha.
 


davehc

Essential Member
Premium Supporter
#7
I agree, Kyle. But there are deeper problems with running Linux, also, than bench testing a vanilla install - I am sure you would agree!
If you read my posts, you might get an impression that I am a 7 "fanboy"
Far from it. I was offered the opportunity to begin early testing of 7, and decided to pass on any information I could glean from my tests and others on the web.
Fact is, I so heavily flamed 7 on another site that a senior moderator threatened to bar me from the site.lol
 


Kyle

New Member
#8
I guess I've become a 7 fanboy of sorts, but maybe that's just 'cause I've been waiting for something like this since Longhorn's alphas hit my desktop. Windows 7 is what I had hoped Vista would be. It's been a long wait, and I'm just glad Redmond didn't f%$^ around too much this time.
 


#9
Your tests are really a little meaningless. The whole point of Vista and, improved, 7 was the availability of resources. This meant loading more DLL's etc at the initial start.
As we move on to updated OS's, and this will include Linux distributions, it is obvious that more demands will be made on hardware.
The memory allocations, in the tests, do not appear to be paramount even with 3Gbs. Xp is, of course, very low, but has poor memory management and will load up to a higher figure when using apps.; Vista was similarly inclined, but handled the memory a litle better. Windows 7, even in Beta, is exrtemely good at handling it. Solution, as we move along to new OS's (Its going to be even more demanding if anyone can afford a touchscreen) Put more ram in.
Hard disk space? Are you serious? When Xp was being planned, most hard disks were really small. Today, the difference in the \"footprint\" is not even a consideration. One point there. Many users in the XP days, slowly had to update their hard disks, as , not the OS, but most modern software began to bloat. One example springs to mind (there are many) Nero.
Fwiw. I just checked my Vista installation against Windows 7. Running the same programs, and with this site open, Vista is using 284mbs and Windows 7 is using 295mbs.

But, maybe we should all look at this. Remember this was an early release, with compatibilty and other issues. Win 7 beta has improved quite a lot since then.
Windows 7 build 6956 vs. Windows XP SP3 | Hardware 2.0 | ZDNet.com
I disagree with your opinion on MORE RESOURCES & FOOTPRINT issues. Many years have past since MSDOS 5 that I started with and issues of memory consumption & space are still big issues since the days of 124 MB HDD. Resources are to be made available, but noway compromizing the performance, otheise why care about fitting more space into smaller chips! And that is where the OPTIMIZATION thing comes in. And space? lol, if you don't care about space issue, you are most probably making more garbages than products.
 


Mike

Windows Forum Admin
Staff member
Premium Supporter
#10
Really the resource usage and footprint has become minuscule in the grand order of things - when we consider for one moment that it will soon be possible to build a desktop PC with 24GB of RAM and a 12 core processor. The limitations of this test are to me that they do not take into account real world scenarios, the variables of which cannot be measured by simple benchmarks to begin with. The problem of course is that we are simply measuring usage and not performance, which I would truly be interested in seeing the release candidate's performance vs. XP in raw numbers on a multitude of tests using different hardware. Of course when an operating system uses less memory this is usually good, but the difference on a modern system with a modern processor between Vista and XP is still virtually negligible; this likely due to kernel-level optimizations for multi-core processors and applications becoming more dependent on multi-core. (Let us not forget too that for single-core applications, the latest processors will now actually increase the multiplier slightly to give a boost). So in this instance we should measure raw performance when it comes to XP vs. Windows 7.
 


#11
I just started using 7 again. I had and earlier beta release and after several BSOD's I uninstalled it. I'm now using 7100 and so far it's OK. :D
 


#12
the way I look at this whole topic of comparing windows XP windows vista and windows 7 it shouldn't be done each O/S is a system of its own not a better version of the one before even though 7 is built on vista's kernel it's still an operating system of it's own so I personally don't compare O/S to each other I use Xp I use Vista I use 7 and I use some LINUX Distros all depends on what I want to do. and to me None are like one another so they are not really comparable.
 


#13
I think one area you did not touch on is the size available today of HD to when XP was around, they now have got ginormus so the need to save space is of less worry and you can have less compressed files to function quicker. Same with Ram. My systems run about the same as you state in your tests whatever the system is, the only problem is the processor speed needed to produce graphics and run todays programs. I think that Win 7 is about the same as XP just trimmed up by fractions of seconds and faster processor speeds give it a slight edge.
 


#14
XP/Vista/7 Tests

I am interested in the following tests:

1. Bootup
2. Shutdown
3. IE8
4. Launch Word Processing
5. Launch Spreadsheet
6. Search for one specific file on disk


Thanks
 


Kyle

New Member
#15
Going to address some issues.
I abandoned this project a while ago, as I realized that there's no real way to test certain things in a VM, and things that ran one way in a previous OS may not run the same in a new one, so technically OS benchmarking is unfair, even if raw numbers are used and it seems there are differences.

Jeremy, most of those things are impossible to test correctly, and I will outline the reasons below.

#3. What are we comparing IE8 to here? We can't 'really' compare it to other brands of browsers, firefox, chrome, etc, as technically they don't conform to the same specifications. It would be the equivalent of comparing a 1967 Mustang to a 2009 Mitsubishi Spyder. When they were designed, they didn't have the same available resources, they were built optimized in different areas, and so forth. While firefox may load google faster, IE may load facebook faster. It's all about optimization and specifications.

Even comparing IE8 to IE7 would be inaccurate, as concerning Windows 7 that would be technically impossible (IE7 can't be installed). If you're talking about cross platform, IE8 installed natively on XP, Vista and 7; again, there's a problem with specification and optimization. I'm betting without modification, on the same system, it would work better on 7 than XP; however, if resources were modified to be equal, XP's lighter OS base would likely push the win for it.

4+5; Depending on the version and make of the Word processor and spread sheet, they may be optimized for different systems, and therefore an unfair advantage would be had to the OS they were optimized for. How do you adjust for that?

6; That is impossible to accurately check, as the same file will not be placed in the same place on the disk by all the OS'. While the XP disk may have that file near the center of the platter the 7 disk may have it on the outside, creating an unfair advantage for one of them.


Bootup and shutdown are, however, valid tests.
Some results I've picked up;

Boot
XP SP3: 42 Seconds
Vista SP2: 38 Seconds
7 RC: 32 Seconds

Shutdown
XP SP3: 24 Seconds
Vista SP2: 19 Seconds
7 RC: 12 Seconds
 


#16
Windows 7 Benchmarks timings

Good work Kyle - thank you

The shutdown and boot timings were great. Windows 7 has a definite advantage over the other OSs - really good info; although if you seldom shutdown/reboot (my wife never shuts down her laptop) it may not matter.

I really appreciate your point on the other tests. It is very difficult to get apples to apples comparisons (no pun intended). However, overall speed of operation with normal functions is really important. My list was just a short list of indicators of speed. For example: if you have 2000+ pictures how long does it take to fully paint the screen with the various OSs/32bit and 64 bit? Any ideas on how to test all of this? In the end 'is Windows 7 worth the time and expenditure' or should we stay where we are (XP or Vista). After all, to change is a big deal, especially if a clean install is recommended (which I support).

Thanks Jeremy
 


#17
Systems themselves have had to play catchup as well. Hardware bought now has more than enough juice for almost 80% of users. The people that always knock new systems, Apple users included, are people that are trying to run new OS under systems that are simply outdated. These things are better to look at outside the box. If we all took this approach we would want windows XP running on our Commodore 64s.
 


Highwayman

Extraordinary Member
#18
Just thought i'd chuck my money in the hat too...

Windows 7 loads to desktop and runs all the default apps etc with a memory footprint 350mb LESS than it did on my Vista x64 sp1 (same software installed), and 127mb LESS then it did using XP Sp3.....so in my boot WIN-7 is best OS of the three and its not even finnished....lol
 


#19
Really the resource usage and footprint has become minuscule in the grand order of things - when we consider for one moment that it will soon be possible to build a desktop PC with 24GB of RAM and a 12 core processor. The limitations of this test are to me that they do not take into account real world scenarios, the variables of which cannot be measured by simple benchmarks to begin with. The problem of course is that we are simply measuring usage and not performance, which I would truly be interested in seeing the release candidate's performance vs. XP in raw numbers on a multitude of tests using different hardware. Of course when an operating system uses less memory this is usually good, but the difference on a modern system with a modern processor between Vista and XP is still virtually negligible; this likely due to kernel-level optimizations for multi-core processors and applications becoming more dependent on multi-core. (Let us not forget too that for single-core applications, the latest processors will now actually increase the multiplier slightly to give a boost). So in this instance we should measure raw performance when it comes to XP vs. Windows 7.
I agree completely Mike.. :)
 


#20
W7 is answer to Vista

Hello,

The meaning of W7 is an answer to Vista, ubuntu is totaly different en XP is an totaly different lokking and working OS.
So wen you look Vista-W7 you see that W7 is a good following of Vista.

Silas
 


This website is not affiliated, owned, or endorsed by Microsoft Corporation. It is a member of the Microsoft Partner Program.