The most obvious ones of course, are the system memory and video card memory. For example, a computer with a slower CPU but equipped with more RAM (random access memory) will load programs just as fast or even more so, than on a computer with a faster CPU but fewer RAM. The same thing with video cards; if it has plenty of memory, loading up and redrawing the screen will be much faster and smoother.

Besides the system memory, the Front Side Bus (FSB) plays a role as well. This is the device that connects the memory with the CPU. So the velocity of the FSB will be crucial in benchmarking CPU speeds too.

Another factor that affects benchmarking CPU speeds is its efficiency in handling assigned tasks. For instance, a 1 MHz (megahertz) CPU can perform a million cycles per second. While a 3 MHz is technically three times faster than the 1 MHz, it will not necessarily accomplish things exactly three times faster because the 1 MHz might be better at handling a task during the cycle.

The cache is also used for benchmarking CPU speeds, but again this can only indicate so much. The larger the cache the quicker the retrieval of the information (i.e., an L2 cache of 1 MB can store more data than an L1 cache of 256 kb). However, you also have to consider at what speeds the cache is working. If it does not make full use of the speed the CPU, then it will work less effectively.

Finally, benchmarking CPU speeds will depend on the tests being conducted. For example, you might see a series of tests performed on CPUs in a magazine or website. But read the manner in which the tests were performed. You will see that they use a variety of applications, from word processors, spreadsheets, 3D applications, photo editing software and Internet browsers. You will see statistics showing which one loads the fastest and slowest, etc.

But are those facts really useful? It is unlikely that you use all the software that were used for the tests, and the way they were run will not be the same way you utilize and load them. Also, it will not escape your attention that the difference between most CPUs is only a few seconds, which you will hardly notice when you are actually working on a project.

Benchmarking CPU speeds can be a useful gauge and tool for new buyers, as it will give them some idea of the velocity of the computer they are going to buy. However it is important that the other factors mentioned above be accounted for too, so you will know exactly what you are getting.