When we go to overclock a memory kit we first toss it into a system that is known to OC well, then we make sure it runs at the standard timings, next we start lowering the timings to see how low we can get with stock voltage and frequency. We go from there increasing frequency, lowering timings, and increasing voltage until we find stable overclocks that we can test. In the case of this kit that was a bit hard. We decided to use an X79 test bed, and we weren't getting anywhere, so we felt like maybe it was hampering us as this kit was designed for Z68 platform. We proceeded to change test beds. We ended up, surprisingly, with the same overclocks. We decided to just use our X79 data, so please be aware that the Z68 platform will have better performance and the same OC.
|HDD||WD 80GB SATA|
|OS||Windows 7 Standard no Tweaks|
When we loaded our X79 system, we loaded optimized defaults and we were surprised that it automatically set to 1600mhz, but with very loose timings.
So then we went and loaded XMP:
Here is the SPD table:
This was our best OC: (1866MHz, at 11-11-11-28 T1 1.65v)
Voltage didn't seem to help at all, except when we went for 1866mhz, where we used 1.65v.
So after we got the highest MHz and we couldn't tighten timings we decided to reduce the frequency to 1333 and see how much we could tighten our timings:
So for our performance tests we will use the following configurations:
1600MHz 11-11-11-24 1T-Stock
1600MHz 9-9-9-24 1T-XMP
1866MHz 11-11-11-28 1T
1333MHz 8-8-8-24 1T
1333MHz with those reduced timings seems to be better than 1600MHz at stock.