In our performance tests we checked out the speed of our LGA1155 platform with different dual-channel memory kits. Each kit was tested in two modes: 1) automatic configuring (every timing is set by the mainboard's BIOS according to the XMP data, the only exception being the HyperX Genesis KHX1600C9D3K2/8G kit which doesn’t support XMP, so we had to manually set it up according to its specs) and 2) highest clock rates and best timings (as achieved in our overclocking tests above) set up manually in the mainboard’s BIOS.
First of all we want to run synthetic benchmarks of memory bandwidth and latency. We will use the MaxxMEM2 suite which can run both in single- and multithreaded mode.
The clock rate of DDR3 SDRAM does affect memory subsystem parameters on Ivy Bridge platforms. For example, increasing the clock rate of the Kingston kits by 533 MHz helps improve the practical bandwidth and latency by 8-10% with single-threaded access. The speed of multi-threaded access grows up even more, by 20-25%.
It seems that overclocking your memory subsystem and purchasing high-speed DDR3 SDRAM makes some sense, but don’t forget that the diagram is based on synthetic benchmarks. We don't usually see such a sharp difference in real-life applications. Anyway, we guess we need to take a look at some real tasks that make a heavy use of system memory.
In some usage scenarios high-speed overclocker-friendly memory can indeed be useful, especially if it isn't much more expensive than ordinary products. Memory subsystem parameters can affect the speed of gaming applications, too.
Of course, a game's frame rate depends on the graphics subsystem performance in the first place. However, memory frequency is an important factor here. When high, it can help you increase your gaming PC's performance by a few percent. So, choosing the right memory is important. And the most important factor about DDR3 SDRAM is its clock rate.