Corsair Dominator GT 3x2GB PC3-15000 Triple Channel Memory Review

Benchmark Methodology

We swear, this is the last time we comment on the lack of XMP profiles. Well, at least until the conclusion anyway. But it is a big part of this kit so it needs to be talked about. The lack of the XMP profiles has required us to manually set the timings for the results in stock form. These settings were derived from the XMP profiles on the memory as best to what we could set given the BIOS options of the DFI. All other BIOS settings are left at default aside from locking the i7 965 processor we are using today to 3.2GHz for this set of results. The grey results are going to mirror these exact clocks and settings but instead of a 3x2GB kit, we will be using a 3x1GB kit. This is a continuation of our last Dominator memory review and wanting to test the difference between 3x2GB and 3x1GB for performance.

The final set of results will be draped - appropriately enough - in red. This color of bars in the graphs will depict our performance at the maximum 6-7-6 and 7-8-7 stable overclocks we achieved with these modules. It isn't actually right at the limit that we found with the Dominator-GT modules for 7-8-7, but knock them down just a notch allowed for a direct comparison between the two timings sets with the same CPU clocks. For all of the benchmarking done today, we will be relying on the services of a wonderful EVGA GTX 295 at its default clocks of 594 / 2052 /1296. Here is how the rest of the test setup will be derived as far as operating system and software goes:

1. Windows Vista x64 w/SP1 is installed using a full format
2. Intel Chipset drivers and accessory hardware drivers (audio, network, GPU) are installed followed by a defragment and a reboot
3. At time of benchmarks the latest drivers were downloaded from their official web sites as the latest drivers, most notable, NVIDIA GeForce Release 180.48 WHQL
4. Programs and games are then installed followed by another defragment
5. Windows updates are then completed installing all available updates followed by a defragment
6. Benchmarks are each ran three times after a clean reboot for every iteration of the benchmark unless otherwise stated, the results are then averaged