CacheBoost
CacheBoost optimizes the System Cache-Management of Windows XP/2000/NT and upcoming Windows .Net Servers. This results in an best performance boost - without any new hardware, any hardware tweak, and without restricting your system's stability. Simply install CacheBoost and you are ready to go. No special configuration or rebooting is needed. Within 5 minutes you activate unused resources of your Windows XP/NT/2000 and Windows .Net systems.

Above image how data is read and written to disk using System Cache
Windows uses system cache which normally improves the speed of accessing files. System cache is part of system memory reserved by windows for its caching needs. The purpose of cache is to store data temporarily to improve performance. When a file is read from disk it is store in system cache. If there is another read request for the same file, windows can retrieves it from the faster system cache memory, instead of reading it again from the slower hard disk. Windows accesses your hard disk for reading and writing through its integrated System Cache Management. Unfortunately, Windows XP/NT/2000 file caching shows a misbehavior when handling file operations. As windows allocates main memory for the cache as and when necessary. After some time there are less and less memory resources available for running applications and bloated cache. If applications request new memory, the system will be forced to swap out currently unused memory areas into the paging file, thus requiring read/write access to the hard disk. This paradox situation leads to a dramatic decrease in system performance, caused by the file caching, which slows down your machine instead of making it faster.
Up to now, the user could neither control windows behavior nor adjust windows cache to its personal needs with the only choice of restarting the system after some time. CacheBoost solves this problem by taking control over the resource management for file caching and applications. CacheBoost is result of two years research on Windows Memory Management. CacheBoost is an Intellisense performance upgrade for Windows XP/2000/NT. Applications can now allocate memory without forcing windows to swap. This leads to a performance increase and to lower response times, especially if you are running many applications at the same time.
Here are the result of our tests with Cache Boost.
Tests Done |
With Cache Boost |
Without CacheBoost |
Zipping of 5000 files ( html and graphics file around 300 MB of data) on Windows 2000 Prof. Intel Celeron 733MHz , 256 MB. |
Time taken 5 minutes 22 seconds.
Free memory before : 144 MB
Free memory after : 140 MB
50% performance boost. |
Time taken 7 minutes 10 seconds.
Free memory before : 142 MB
Free memory after : 96 MB |
Windows 2000 Server running continuously for one week serving 25 clients.
Server dual Intel P-III, 256 MB. |
Free memory between 64 MB to 84 MB with average 35% swap file usage. Server after one week continuous running was still giving just booted feel. |
Free memory between 27 MB to 45 MB with average 65% swap file usage. Server was restarted two times in week. |
Windows XP Home Edition. Movie maker was used to create a movie from 25 Picture Clips.
(Scanned pictures from family album 640x480 - true color.)
Intel P-III 1 GHz , 256 MB. |
Time taken 4 minutes 41 seconds.
Free memory before : 125 MB
Free memory after : 126 MB
54% performance boost. |
Time taken 6 minutes 25 seconds.
Free memory before : 115 MB
Free memory after : 72 MB |
| Windows 2000 Professional on Office PC ( Intel P-II 450 MHz 256 MB Memory). Application running Microsoft Word, Outlook, Internet Explorer and Excel. Time period was 9:30 AM in morning and 5:00 P.M. in evening . |
Free memory was constant in between 125 MB and 140 MB. Windows Page file average usage was about 18% with lowest to 15% at System start up and peek was 22%. In evening system was giving just booted feel. |
Free memory was fluctuating in between 125 MB and 84 MB. Windows Page file average usage was about 35% with lowest to 17% at System start up and peek was 50%. |
|