[display_podcast]
Windows Vista ships with a technology called ReadyBoost that is intended to increase the performance on systems that aren�t equipped with an adequate amount of RAM. The idea is good, the user simply plugs in a USB key and lets Vista use a part of the key�s capacity as a sort of �extra RAM�. No need to run out and buy expensive (which they really aren�t anymore) RAM modules; all you need is a fairly fast USB key. The setup is really easy; if Vista recognizes the USB key as being ReadyBoost capable when the key is plugged in, Vista asks the user if he or she wants to increase the system's performance with ReadyBoost. If the user clicks yes, Vista will ask how much of the USB key�s capacity should be used for this and after the user has specified the amount and clicked OK, the setup is done.
Now it�s time for a history recap. Does anybody remember some 12 years ago when software like RAM Doubler and others were quite popular? They claimed that through various ingenious techniques they could make the system manage the memory so much more efficiently that installing the software would create a performance increase equal to doubling the amount of physical RAM in the system. This, of course, is also a good idea.
However, what prevents the historical software approach and today�s ReadyBoost to go to history as great ideas is that they suffer from a minor flaw. They don�t work! Microsoft claims that �Windows ReadyBoost relies on the intelligent memory management of
Windows SuperFetch and can significantly improve system responsiveness.� The keyword in this quote is �can�; they don�t say that it will. Now, I don�t doubt that ReadyBoost relies on the intelligent memory management as they say; it�s just that so far I haven�t seen it improve the performance in any noticeable way on the systems I tested; significantly or otherwise.
I admit that my tests weren�t conducted in a very scientific manner. I booted into Vista, performed a couple of operations, such as resizing a hundred images in Photoshop, and I used my watch to measure how long it took to complete the operation. Afterwards I plugged in a fairly fast 4GB Sandisk Cruzer key, assigned the recommended portion of it to ReadyBoost and re-ran my tests. My test system was a regular Lenovo desktop system and I ran the tests with 1GB, 512MB and 256 MB of RAM installed. In addition to this, I also tested on a Lenovo ThinkPad T60 with 1GB RAM. I couldn�t see any improved performance in either of my tests. For instance, in my desktop test with 512 MB RAM installed the resize operation took 79 seconds without ReadyBoost and 77 with; obviously I would never have noticed the difference without a watch in my hand. I did the 256 MB test just for fun and it was a long time since I�ve seen a hard drive swap like that. Again though, I saw no difference in the test with ReadyBoost, it swapped just as much as without and the operations took just as long to complete.
To conclude, my tests were pretty basic and if anybody has an idea how to make ReadyBoost work better, I�d be happy to listen. Meanwhile, I�m not going to use ReadyBoost on any of my systems, because the minute it takes to set it up has proven to be too difficult to get back. For Vista users looking for a true performance boost, I recommend installing XP. Some would call that a down grade, but I�m not too sure.
