Legit Reviews

To start with I chose not to adjust the BCLK on my unlocked i5 2500k so that any tests I ran would represent the pure gain from over clocking the memory.  Leaving the BCLK at 100 did potentially limit the upper end frequency but I was more than happy with my results.

First I checked to see how tight of timings I could get just running at the rated 1600Mhz and found I could drop my CAS to 5 for a 5-7-8-19 but when I ran a few benchmarks it did not make that big of an improvement over the rated XMP profile of 7-8-7-20.

I started to push for higher frequency and was able to get a perfectly respectable 1867Mhz at 6-8-8-20 on a 1T CR.  This was a nice 17% increase over the rated XMP profile of 1600Mhz with decent timings.  I have found that really only increasing frequency and lowering CAS have much affect on performance with today's well made DRAM so this is where I focus.  I usually start off with all the other timing pretty lose until I get the CAS down and frequency up and stable.  Then I work on tightening down the rest.

Knowing this was Dominator memory ,even though this wasn't the speediest of the Dominator line, I knew it had to at least perform as well as the Vengeance series tested here on Legit Reviews back in Feburary so I pushed further.  I kept to a maximum of 1.65v as I did not want to push my test rig out of spec. but I have read that some have gone as high a 1.7v on their DDR3.  This has the potential of harming your Sandy Bridge CPU so I left that to the extreme team.

The best I could get without touching the BCLK was 2132Mhz with 8-10-9-24 timings.  This is 33% higher than the rated speed of 1600Mhz. 

What does a 1/3rd more frequency do for your game play and encoding, jump to the test page to find out!