NVIDIA's 790i chipset-based motherboards are widely recognised by enthusiasts as being among the best when it comes to high frequency DDR3 overclocking. Providing both the ability to run memory modules in sync, on a divider or 'unlinked' from the main CPU bus speed, they are the ideal platform for testing out the EPP2.0 enabled Xtune kit on review today. This, combined with the rest of the hardware listed below, will ensure that no other part of the system acts as a bottleneck:
Intel Core 2 Quad Q6600 "G0" 2.4GHZ 2x4MB
ASUS Striker II Extreme 790i
Aeneon Xtune DDR3-1866 4GB
ASUS Radeon HD 4870x2
ATI Catalyst 8.6.64789
Windows Vista Ultimate SP1 + Updates
To guarantee a broad range of results, the following benchmark utilities will be used:
Synthetic Benchmarks Overclocking
• Lavalys Everest 4.0
• SuperPI Mod 1.4
• Enemy Territory: Quake Wars
While our natural-born instinct as overclockers was to jump right in and start pumping extra voltage through the Aeneon kit, there is one important issue that needs to be explored first: stability. Too often here at Overclock3D, we've received memory kits that simply don't work at their advertised speeds or voltages, leading to system crashes, corrupt OS installs and a lot of wasted time. Therefore, our first test was to see if the Xtune modules would work flawlessly straight out of the box...
DDR3-1866 / 10-10-10-30 / 1.5v
With a stock voltage of 1.5v and claims from Aeneon that "No Overvoltage Is Required", this immediately set alarm bells ringing. After all, most other high performance DDR3 kits we've tested recently have required at least 1.8v in order to function correctly. However, after inserting the modules, enabling EPP2.0 in the BIOS and setting the memory voltage to 1.5v, the system booted first time at DDR3-1866 and effortlessly passed a full 30 minutes of OCCT stability testing along with a series of benchmarks. Amazing work Aeneon!
DDR3-1974 / 10-10-10-30 / 1.5v
The next test was to see just how far we could overclock the modules without making any adjustments to the voltage. Already impressed by the kit's ability to run at DDR3-1866 with only 1.5v, we certainly weren't expecting to see much in the way of overclocking at such a low voltage level. Boy were we wrong...
After disabling EPP2.0, which had us locked at DDR3-1866, and dropping the CPU multiplier to 8x so that we could push the FSB higher without putting too much strain on our CPU, the maximum memory speed we managed to reach with 100% stability was DDR3-1974. Interestingly, the modules also booted and were 'desktop stable' all the way up to DDR3-2000, but no amount of voltage could stabilise this enough for a 3DMark run. However, a little extra voltage did help to get this result...
DDR3-1994 / 10-10-10-30 / 1.56v
So close! Only 6mhz shy of DDR3-2000 at 1.56v. Unfortunately, the Xtune kit didn't scale any better with the voltage set beyond this, which is a shame because it could have easily had Micron's famous D9-based modules on their hands and knees begging for mercy. But with all said and done, going from 1866MHz to 1994MHz with only 0.6v is a damn good result.