There are many Internet sites recommending to disable services to optimize the performance of Windows XP but it is extremely difficult to find tests before / after. In this article we are going to undergo some tests PCMark '05 to a Dell Precision M90 features an Intel Core 2 Duo T7200 2 GHz and 1 GB of RAM. This notebook is not a lightning war the effects of optimization should be more obvious than on a fast machine.
Windows XP SP3 is freshly installed and all updates to date have been conducted, including Windows Search 4. No other application is installed on the workstation.
At startup, without modification, from 183 to 193 MB of RAM are used.
I then followed the setup guide services Black Viper using Safe configuration. Over a few reboots, the RAM usage was 182 to 195 MB, which is equivalent to the non-optimized configuration.
Here now the PCMark '05 test results before and after optimization:
| Before | After | |
|---|---|---|
| HDD - XP Startup | 6823 MB / s | 6720 MB / s |
| Physics and 3D | 170,051 fps | 170,561 fps |
| 2D - Transparent Windows | 890,024 windows / s | 883,384 windows / s |
| 3D - Pixel Shader | 220,745 fps | 218,047 fps |
| Web Page Rendering | 2,614 pages / s | 2,619 pages / s |
| File Decryption | 54,122 MB / s | 53,676 MB / s |
| 2D - Graphics Memory - 64 lines | 1506.480 fps | 1508.011 fps |
| HDD - General Usage | 4580 MB / s | 4571 MB / s |
| Multithreaded Test 2 - Text Edit | 133.48 Pages / s | 129,737 pages / s |
| Multithreaded Test 2 - Image Decompression | 26.96 Megapixel / s | 27895 Mpixels / s |
| Multithreaded Test 3 - File Compression | 7377 MB / s | 4689 MB / s |
| Multithreaded Test 3 - File Encryption | 20,990 MB / s | 26,260 MB / s |
| Multithreaded Test 3 - HDD Virus Scan | 27,454 MB / s | 35,160 MB / s |
| Multithreaded Test 3 - Memory Latency - Random 16 MB | 8011 MAccesses / s | 8207 MAccesses / s |
| Before | After | |
|---|---|---|
| File Compression | 8748 MB / s | 8722 MB / s |
| File Decompression | 131,098 MB / s | 129,379 MB / s |
| File Encryption | 53,473 MB / s | 53,807 MB / s |
| File Decryption | 54,104 MB / s | 53,916 MB / s |
| Image Decompression | 27,521 megapixel / s | 27,792 megapixel / s |
| Audio Compression | 2512.249 KB / s | 2492.050 KB / s |
| Multithreaded Test 1 - File Compression | 8641 MB / s | 8887 MB / s |
| Multithreaded Test 1 - File Encryption | 54,136 MB / s | 52,624 MB / s |
| Multithreaded Test 2 - File Decompression | 64,168 MB / s | 66,647 MB / s |
| Multithreaded Test 2 - File Decryption | 27,626 MB / s | 27,350 MB / s |
| Multithreaded Test 2 - Audio Decompression | 1009.103 KB / s | 1027.911 KB / s |
| Multithreaded Test 2 - Image Decompression | 14,194 megapixel / s | 14,210 megapixel / s |
One can see that both configurations are equivalent globally. Sometimes the configuration after optimization is a little slower, and sometimes a bit faster. This remains in my view entirely within the margin of error of the test and this is also checked by repeating the test several times to turn: the results vary in one way or the other, without special logic.
To conclude, I think it is unnecessary to make (or lose) time to disable Windows services in an attempt to optimize a business machine, especially since this could cause problems later. Who will remember that this service was disabled and that he just needed to run a special application just as one meets frequently in the professional world?
Tags: Optimization , Windows XP
