TEST BED
Our primary test rig based on the Intel Core i7-3930k Rev.2 CPU and the Gigabyte X79-UD7 Mainboard will be used for this review.
TESTING METHODOLOGY
Whenever i test something which can be measured in terms of performance i always like comparing it with similar products just to see how well it stands its ground against them. However since this is the first Quad Channel kit i will be reviewing i can only compared it against Triple Channel kits which i have here with me. Ordinarily i would skip doing so since we are talking about different platforms/systems all together (LGA1366/LGA2011) but i do want to showcase how the latest Quad Channel technology compares to the previous one. So today i will be comparing the latest Kingston HyperX Genesis 16GB 2133MHz kit with the HyperX T1 6GB 2000MHz and the HyperX 12GB 2250MHz kits. Also since the XMP profile of the HyperX Genesis kit has a default command rate of 2 (the other two kits are running at 1T) i will also be performing tests with 1T to see just what more you safely can squeeze out of it in terms of performance without much effort and without increasing the voltage.
To give you the best and most accurate results possible we decided on using several methods including both benchmarking programs/suites and games. The benchmarking programs used are the well-known AID64 Extreme Engineer 2.3 version (formerly known as Everest Ultimate), the Sisoftware Sandra Pro 2012 version, the old but quite reliable ScienceMark 64Bit and the all-time classic SuperPi 1.5. Finally we needed a game with lots of fps (simply put) to better showcase and point any differences between kits so from now on we will be using a HAWX II timed demo in this kind of reviews. Every test was performed a total of 4 times after which the average result numbers were recorded into the charts. For all tests both CPUs (920/3930k) were overclocked at 4GHz in order to keep the results as accurate and fair as possible while the graphics card used in both systems is the Gainward GeForce GTX 590 Golden Sample.