Benchmarking Methodology:Here we go with another round of benchmarks. The last 2x2GB DDR3 memory we had come through was a little disappointing on behalf of the motherboard issues we encountered. That won't be the case today as we clearly saw in the overclocking section. We have a full slate of possible configurations to bench and compare and we intend to spread the spectrum. In the past we found that simply changing memory frequency on the 790i through un-linked memory left some rather un-desirable results. So today we will be running synced as much as possible.
As always, the grey bar in the graphs (1) will indicate the rated frequency and timings of the Mushkin Ascent 2x2GB kit of memory with the system at bone stock settings for the E8400 resulting in a 3GHz CPU. These grey results will be a result of us manually adjusting the main timings since there is no EPP profile to take care of that for us. The secondary timings will be left at the BIOS defaults. Essentially, this is what the basic user of this memory might have setup as all it takes is a couple minutes to do so with no in-depth adjustments being made or tested.
The middle blue results in the graphs, (2), will portray the performance of the Mushkin Ascents at the same CPU frequency, FSB, and will both be run at 2T. These two settings are essentially the same as our overclocked maximums we found on the previous page. The 6-6-5 timing set is run synced on the 5:4 ratio while the 7-7-7 results were run un-linked to almost match the highest overclock of 910MHz with that timing set.
The last red result in the graphs (3) is used to identify the "hottest" overclock we had at 8-8-8. In our overclocking we were only able to get 990MHz stable but wanted to stay synced at the same FSB as the other two overclocked results so we are working at 1000MHz which isn't 24/7 stable but more than stable enough for the benchmarks we are going to be running today. The 10MHz difference isn't much and will provide results very similar to what our maximum overclock would churn out.
For all of the benchmarks, appropriate lengths are taken to ensure an equal comparison through methodical setup, installation, and testing. The following outlines our testing methodology:
a/ Windows is installed using a full format.
b/ NVIDIA chipset drivers and accessory hardware drivers (audio, network, GPU) are installed followed by a defragment and a reboot.
c/ Programs and games are then installed followed by another defragment.
d/ Windows updates are then completed installing all available updates followed by a defragment.
e/ Benchmarks are each ran three times after a clean reboot for every iteration of the benchmark unless otherwise stated, the results are then averaged.
We have also listed the various benchmark versions as they change on a weekly basis and sometimes different versions can result in different outcomes. All benchmarks are ran a total of three times and averaged out with results posted to the same decimal point that the results are in. The one exception to this is the UT3 benching as the results can vary a fair amount from run to run so we do ten runs to help alleviate the variance a little bit. |