Friday, October 03, 2008

Vista ReadyBoost & SuperFetch

Another one of those funny-sounding technologies that use a quite new concept and promise this and that. Well, to find out if they really work sometimes is enough just to browse the net and read the reviews.
From what I could find it seems that ReadyBoost is not really an option since the system already has 3GB of RAM.
This exercise went well and just saved me $40 on a new flash drive that I might have used to test ReadyBoost (and find it didn't have any visible effects on the system).

There is also another technology - SuperFetch. There is an interesting review on Tom's Hardware, explaining some background technologies, like caching, memory usage, etc. SuperFetch apparently works better than in pre-caching in WindowsXP. ReadyBoost is actually used as a cache storage for SuperFetch. This draws a conclusion that the improvements in performance are to be felt with commonly used programs. All the tests that just run common speed/throughput tests on a system would not see any performance improvement as SuperFetch had not had the time to analyze application usage during the test runs.
The real difference comes from separating the system drive, containing the swap file, and data files. These data files get cached in the flash drive, using ReadyBoost.
Test results mimicking standard computer usage, and not running any standardized performance tests, show there is significant improvement. But, on older systems, that were under stress by just installing Vista, anyway. On systems with 1GB of RAM or more, the results are not that dramatic. In practical terms, almost non-existent.

So, the conclusion? All these technologies are nice to have, of course. Stick to the specifications and use new operating systems on new hardware. Give them the recommended configuration or more and don't expect miracles. :)

No comments: