What do we look for when we buy memory? RAM can be measured in nearly a million different ways, and the effect of all of these numbers is shrouded in secrecy, until now. GotFrag Hardware has recently acquired a Corsair XMS 3500 2x1GB memory kit from the very nice folks at Corsair to review. The review for this product will be coming shortly, but in the mean time, we have assembled this article. In it we will demonstrate just how all the numbers actually impact overall FPS, and we'll also add our two cents about whether investing in that particular specification would be smart or not.

We have tested this Corsair RAM at different timings, channels, sizes, and command rates. At the end of these tests, we hope to dispel or confirm the many myths that are commonly taken as fact when considering the purchase of RAM.

To properly study the effect of the RAM, we must make sure that it is an FPS limiting factor. FPS is limited in two ways: the power of the 3D graphics card, and the power of the rest of the system. This is referred to as GPU and CPU limiting. For more information on this, visit our earlier article HERE. Using that article, we have determined to use Quake 4, CS 1.6, CS:S, and HL:DM in our tests. All tests were done at high quality, running at 800x600 resolution. At this setting we are guaranteed to have our FPS limited by something in the rest of the system as opposed to our 7800GT. The rest of the setup is as follows:

• AMD Athlon64 X2 4200+
• Asus A8N SLI Premium
• EVGA 7800GT
• 1.16TB Raid5 array