Overall performance comparison between different computers under Windows XP

[Multibooter]

On 9/14/2024 at 11:37 AM, NotHereToPlayGames said:

How do you spell "too shay"?  

Maybe "tsu schee" ?

In my posting of 10Sep2024 I had displayed a screen shot of the test results of PCMark04 on an Inspiron 7500 laptop (650MHz Pentium 3 SSE-only), run with Windows Media Player 11 installed. The overall score was 568.

I have subsequently restored the WinXP partition to a clean, pre-PCMark partition backup. I then installed Windows Media Encoder 9 and PCMark04 v1.3.0 and then ran PCMark04, i.e. with the default Windows Media Player 9, which comes with WinXP, instead of Windows Media Player 11. The overall score was 617.

 

  

 

The better test results with Windows Media Player 9 show that the last version for WinXP (i.e. v11) is NOT the best version for every computer, e.g. the old Inspiron 7500. The test with PCMark04 shows that installing the system component Windows Media Player 11 will probably degrade the performance of the old Inspiron 7500. Unfortunately PCMark04, in contrast to PCMark05, does not create a log file showing how long the tests took.

The above comparison of Windows Media Player 9 vs 11 with PCMark04 has, however, a major issue: the initial test of 10Sep2024 was not made after a partition restore, so other issues may have contributed to the worse performance of Windows Media Player 11.

BTW, Windows Media Player 11 canNOT be uninstalled, only a rollback can be made. I made a rollback to the previous version (i.e. v9) in two steps:
-> Start -> Run -> enter %windir%\$NtUninstallwmp11$\spuninst\spuninst.exe
-> Start -> Run -> enter %windir%\$NtUninstallWMFDist11$\spuninst\spuninst.exe
I had installed WMP11 NOT by running wmp11-windowsxp-x86-enu.exe, but by first extracting wmp11-windowsxp-x86-enu.exe and then running first the extracted wmfdist11.exe and then the exptracted wmp11.exe

I don't know how long the test with PCMark04 took on the Inspiron, PCMark04 does not create a .log file. The test with PCMark05 took 7:40hrs

PCMark04 does have a special use with very old computers, e.g. the Inspiron 7500 laptop, identifying software and hardware components which slow down the computer, a critical issue for the usefulness of old computers.

One benefit of installing older PCMark04 instead of PCMark05 is that PCMark04 works OK with Windows Classic. To restore my previous Windows Classics, e.g. after uninstalling PCMark05, I made the following 3 steps:
1) -> right-click on desktop -> Properties -> in Themes tab: -> select Windows Classic
2) in a Windows Explorer window: -> Tools -> Folder Options -> select Use Windows Classic folders
 (had been changed to Show common tasks in folders)
 AND: ->select Open each folder in its own window
3) to restore colors used for high-lighting, etc:
 -> right-click on desktop -> Properties -> in tab Appearance:
  - in drop-down box Color Scheme: -> select Windows Classic
  - in drop-down box Windows and buttons: -> select Windows Classic style

I am not sure whether I will keep PCMark04 on the old Inspiron 7500. The 3D test component of PCMark04 still takes many hours. eventually I will test-install PCMark2002. Maybe this old Win98 version is the best version for the Inspiron 7500 under WinXP.

Edited September 15 by Multibooter

[NotHereToPlayGames]

50 minutes ago, Multibooter said:

The better test results with Windows Media Player 9 show that the last version for WinXP (i.e. v11) is NOT the best version for every computer, e.g. the old Inspiron 7500.

Nice writeup!

I'm kind of not surprised that installing "all" WU suggested-as-critical updates actually SLOWS DOWN the computer.  We used to have some XP x86 SP2 folks that swore the same for x86's SP3.

I do know that a slipstreamed x86 SP3 performs much better than an XP that started life brand new, then was updated to SP1, then was updated to SP2, then was updated to SP3, and all of the hotfixes between or after each "service pack".

Kind of wish the ol' x86 SP2 members were still around so that we could see these scores on x86 SP2.

Edited September 15 by NotHereToPlayGames

[AstroSkipper]

2 hours ago, Multibooter said:

In my posting of 10Sep2024 I had displayed a screen shot of the test results of PCMark04 on an Inspiron 7500 laptop (650MHz Pentium 3 SSE-only), run with Windows Media Player 11 installed. The overall score was 568.

I have subsequently restored the WinXP partition to a clean, pre-PCMark partition backup. I then installed Windows Media Encoder 9 and PCMark04 v1.3.0 and then ran PCMark04, i.e. with the default Windows Media Player 9, which comes with WinXP, instead of Windows Media Player 11. The overall score was 617.

 

  

 

The better test results with Windows Media Player 9 show that the last version for WinXP (i.e. v11) is NOT the best version for every computer, e.g. the old Inspiron 7500. The test with PCMark04 shows that installing the system component Windows Media Player 11 will probably degrade the performance of the old Inspiron 7500. Unfortunately PCMark04, in contrast to PCMark05, does not create a log file showing how long the tests took.

The above comparison of Windows Media Player 9 vs 11 with PCMark04 has, however, a major issue: the initial test of 10Sep2024 was not made after a partition restore, so other issues may have contributed to the worse performance of Windows Media Player 11.

As you surely know, in the software requirements of PCMark04, you can find Windows Media Player listed in the version 9 only. The slightly worse performance with Windows Media Player 11 may well be due to the fact that PCMark04 has not been adapted to this newer WMP version. Apart from that, your measured performance difference is only marginal. I have carried out several tests, and the results were never the same.

[Multibooter]

5 hours ago, AstroSkipper said:

As you surely know, in the software requirements of PCMark04, you can find Windows Media Player listed in the version 9 only. The slightly worse performance with Windows Media Player 11 may well be due to the fact that PCMark04 has not been adapted to this newer WMP version.

Maybe, maybe not.

PCMark uses, for example, qasf.dll [=DirectSHow ASF Support] of the Windows Media Player to calculate graphic test values. It would be logical to expect that the use of different versions of qasf.dll, contained in different versions of WMP, would result in different graphic test results. But how could audio-/video-related dlls affect CPU Scores (-8.1%), Memory scores (-10.2%), Graphics scores (-16.1%) and HDD scores (-6.6%) when the test was made with Windows Media Player 11 instead of WMP 9? Big puzzle, but WMP is a Windows component.

Again, I had stated in my posting "The above comparison of Windows Media Player 9 vs 11 with PCMark04 has, however, a major issue: the initial test of 10Sep2024 was not made after a partition restore, so other issues may have contributed to the worse performance of Windows Media Player 11." Maybe repeating the WMP 11 vs 9 benchmark test after identical partition restores could create clarity, but it's too time-consuming for me.

The worse test results with Windows Media Player 11 just don't give me a reason for upgrading from v9 on the old Inspiron 7500, especially since I do not use Windows Media Player, except with PCMark. The apparent worse performance of WMP11, and the consequent rejection of PCMark05, which requires WMP10 or 11, is relevant for my weak, old Inspiron 7500, but should be less relevant for my stronger desktops, their performance and speed is not that important.

Edited September 15 by Multibooter

[Multibooter]

4 hours ago, AstroSkipper said:

your measured performance difference is only marginal. I have carried out several tests, and the results were never the same.

Yes, test results of PCMark do vary. But 6-16% worse is a little high.

Edited September 15 by Multibooter

[NotHereToPlayGames]

13 hours ago, Multibooter said:

But 6-16% worse is a little high.

Agreed.  I have not ran the test ten times and graphed on a scatter plot, but I would expect ten same-hardware runs to all fall well below 6 to 16 percent variation.

[Klemper]

Time of the day, ambient temps are all well known to affect such tests.

[Klemper]

On 9/15/2024 at 12:32 PM, NotHereToPlayGames said:

Nice writeup!

I'm kind of not surprised that installing "all" WU suggested-as-critical updates actually SLOWS DOWN the computer.  We used to have some XP x86 SP2 folks that swore the same for x86's SP3.

I do know that a slipstreamed x86 SP3 performs much better than an XP that started life brand new, then was updated to SP1, then was updated to SP2, then was updated to SP3, and all of the hotfixes between or after each "service pack".

Kind of wish the ol' x86 SP2 members were still around so that we could see these scores on x86 SP2.

XP was gimped down with SP2, it's well known. My snappy 14 inches Celeron notebook was brought down to its knees with SP2 in 2004.

*I had to revert and reinstall the OS from the supplied OEM CD.

Edited September 16 by Klemper
*

[Klemper]

And let me ask, what features SP2 and SP3 brought? Right, except the new "security enhancements", they only removed them.

Service Pack 2 and Service Pack 3 also removed features from Windows XP, including support for TCP half-open connections[52] and the address bar on the taskbar.[53]

https://en.wikipedia.org/wiki/Windows_XP

Edited September 16 by Klemper

[Multibooter]

On 9/15/2024 at 12:32 PM, NotHereToPlayGames said:

Kind of wish the ol' x86 SP2 members were still around so that we could see these scores on x86 SP2.

I am still one of them. My "progressive" WinXP opsys/partition backups are based on WinXP SP2, on computers where Win98 is also installed on another partition. By "progressive" opsys backup I mean a clean previous opsys backup plus clean re-installs of programs which I want to add permanently. For creating a new "progressive" opsys backup I restore the previous "progressive" opsys backup, then make a clean install of new useful software to be added plus detect useful new hardware, then created the new "progressive" opsys/partition backup.

Since about 2017, I usually install the WinXP SP3 Service Pack update, as a 2nd step, after having restored an opsys backup. I have archived my old "progressive" partition backups, specific to each computer, starting in June 2009. For example, I could restore the WinXP opsys backup "WinXP_Inspiron_11Sep2010.gho", so that WinXP on the Inspiron 7500 is back to how it was on 11Sep2010.

I also keep an Install Log .txt file, documenting what software etc was added in each "progressive" opsys backup. With Ghost Explorer I can easily see and extract files contained in the .gho files.

My last clean WinXP opsys backup of the Inspiron 7500 laptop (Pentium 3 SSE-only) was made on 6Jan2024 and contains WinXP SP2, not SP3. and the last clean WinXP opsys backup with the Asus P5PE-VM desktop (Pentium Duo E2200), also with SP2, was made on 11Feb2024. So it would be possible to benchmark these two computers with and without SP3, if I just had the time.

Different benchmark results of SP2 vs SP3 on the Asus P5PE-VM desktop (Pentium Duo E2200) are of lesser importance to me, unless the potential deterioration caused by SP3 would be so serious as to reduce the usefulness of computer.

With the old Inspiron 7500 laptop (Pentium 3 SSE-only), however, a benchmark SP2 vs SP3 might be quite useful to me, especially for deciding whether to add a new operating system "Windows XP SP3", containing software which requires SP3. I am still using System Commander v9.04 as boot manager, on computers where Win98 is also installed, and most of these computers contain a rarely used, additional 2nd instance of WinXP, installed onto an NTFS partition, so there would be a drive letter available for a "Windows XP SP3" partition. BTW my main WinXP is installed onto a FAT32 partition for compatibility with Win98.

I haven't noticed a major performance deterioration caused by the WinXP Service Pack 3 update on the Inspiron 7500. But maybe I just didn't notice it because the old Inspiron 7500 laptop is already sooooo slow, especially with web browsers, so that any additional sluggishness would slip my attention.

BTW my computers become perceptably crisper after a clean opsys restore. Maybe because no junk is restored, maybe because Ghost restores a pretty much defragged partition.

Edited September 16 by Multibooter

[Multibooter]

8 hours ago, Klemper said:

And let me ask, what features SP2 and SP3 brought? Right, except the new "security enhancements", they only removed them.

Service Pack 2 and Service Pack 3 also removed features from Windows XP, including support for TCP half-open connections[52] and the address bar on the taskbar.[53]

https://en.wikipedia.org/wiki/Windows_XP

Some versions of software require SP3 and don't install or run under SP2, e.g. Beyond Compare 4, MiniTool v10, Paragon Hard Disk Manager 15, HDDScan v4.0, MyPal68 v68.13.8, LibreOffice v5.4.7.2

TCP-IP patch v2.23d by LVLLord may resolve the TCP-IP issue http://www.lvllord.de/?lang=en&url=downloads

The removal of features ("security enhancements") was perhaps intended to make it harder for certain download programs, for example when eMule is running 24/7 and the internet provider disconnects and reconnects the internet connection regularly, e.g. every 24 hrs. The patch seems to allow eMule and uTorrent to resume downloading as if nothing had happened. A value of 50 is recommended at that website.

Edited September 16 by Multibooter

[Saxon]

On 9/7/2024 at 2:35 PM, George King said:

@AstroSkipper I have just created WinRAR SFX Repack for 3DMark06. It installs silently and key is automatically added into registry. So you got full version without any additional steps.

 

https://mega.nz/file/c2UylCKB#aNtZJ2hIqniYJiLaL1wQNK6vbI6h1tgLwcopsLJyJXY

High and thank you! Is there a portable DX9 pack exists as well?

[George King]

9 hours ago, Saxon said:

High and thank you! Is there a portable DX9 pack exists as well?

You mean silent DirectX RePack, right?

https://forums.mydigitallife.net/threads/repack-directx-end-user-runtime-june-2010.84785/

 

Just use /ai or /y switches to make it Unattended.

[UCyborg]

Installed all those service packs back then. Or more exactly went from fresh install from CD that got it slipstreamed. Must have re-installed Windows million times back then. Never ran benchmarks. XP always seemed quirky. I remember one time Explorer opening folders with delay after fresh install. Or was that those old Maxtors, what do I know.

Edited September 17 by UCyborg

[Saxon]

23 hours ago, George King said:

You mean silent DirectX RePack, right?

 

https://forums.mydigitallife.net/threads/repack-directx-end-user-runtime-june-2010.84785/

 

 

Just use /ai or /y switches to make it Unattended.

High and thank you! I meant a portable of DX9 pack that doesn't install into the system, only locally, to get this testing programme work.

I don't own any DX9 games, I don't play them.