Is faster RAM really worth the extra cash ?
When you build computers and purchase the peripherals by yourself you have to make several decisions. Will it be an Intel or AMD cpu for example, which hard drive will you choose from the hundreds available, which video card and how much RAM. The newest hardware is normally the fastest but most expensive one and it is always a question if the extra cash that you had hand over is worth the investment performance wise. Would you pay an extra $100 for a speed gain of 1%? Probably not but for 10% or more? The big problem is that you do not really now how much performance you gain when you select more expensive hardware.
Lets take computer memory as an example. RAM, Random Access Memory, puzzles a lot of buyers. There is DDR, DDR2 and DDR3 RAM, different clock speeds and those cryptic CL-5-5-12 (random numbers for visualization) information. The question if a buyer should go with DDR2-800 or DDR2-1066, DDR3-800 or DDR3-1066 with different CL settings is hard to answer. And that's why Tom's Hardware decided to test the available RAM and see how faster RAM affected the performance of a computer.
The results are somewhat disappointing for the faster is better crowd. The fastest RAM tested, the DDR2-1066 4-4-4-12 increased the performance of the system by roughly 10% over the slowest DDR2-667 5-5-5-12 RAM. And that was only in one game, all other tested games saw an increase of only 4%. I think it's interesting to note that faster RAM is obviously not worth the extra cash.
The guide fails to list pricing information which would really be interesting in the context so that it would become clearer how much you'd have to pay for to get the 4% speed increase if you'd pick the fastest RAM instead of the slowest.
Update: While the test is certainly outdated, it still highlights that the performance gain of faster RAM, even nowadays, is usually not worth the extra price you have to pay.