Corsair Dominator TWIN2X2048-8500C5DF Memory

Corsair has been in the memory business since 1994, and over that time have become the most well known memory manufacturer to everyone from the budding geek to the Alpha-Male enthusiast. Always a leader in performance, and innovation, Corsair is constantly looking for ways to make the best better, and stay one step ahead in an increasingly competitive market. 

Today HardwareLogic looks at the high-performance world of Corsair’s newest ass kicker....the Dominator Series memory. Featuring kits from XMS2-8500 all the way up to XMS2-10000, the Dominator Series is not for the faint of heart. If you're in the market for some extreme memory, sit back and follow along as we look at the introductory kit, the Corsair Dominator TWIN2X2048-8500C5DF.

Featuring DHX (Dual Path Heat Xchange) heatspreader technology, our test sample brings a full 2048 Mb of memory goodness to the test bench in the form of 2x1024 modules. Boasting EPP (Extended Performance Profiles) and SLI Ready labeling, these modules are rated at 5-5-5-15 at 2T, 1066Mhz, and 2.2V. Typical ram modules measure around 1 3/16” for height, but these Dominator modules with the DHX heatspreaders checks in at 2 1/16”.

Once you add the Dominator Airflow Fan apparatus (referred to as the DAF from here on),  you’re looking at a total height of 3 3/8”. In our test bed today, there’s zero clearance issues, even though we have a CNPS9500 for a CPU cooler and an EVGA 8800GTS 640Mb videocard installed. Those using large coolers, like the Cooler Master GeminII will be hard pressed to use the DAF as part of their system.

The retention mechanism for the DAF is installed by sliding it over the clips that lock your ram into the motherboard slots. While quite secure, this retention can cause a tad of frustration when working inside a case.

Included in kits with the "DF" suffix, or available for purchase separately,  the Dominator Airflow kit is simply a small aluminum housing which holds three 40mm fans, which improve airflow over the modules, and thus improve stability and performance.

The waspish look of these memory modules translates into stinging  performance in all areas. From a half speed of 533FSB to an impressive 15.7% overclock, these modules performed extremely well. At one point, we hit a 1175FSB, but this failed Memtest after 30 minutes. We could run benchmarks at 1175FSB, but stability  is paramount when dealing with RAM. Subtle failures such as detected in Memtest at 1175FSB can mean data corruption over time and a dreaded OS reinstall. Running these bad boys at a 1157FSB is certainly nothing to scoff at though, and this is what we used for our maximum overclock benchmarks.

The Dominator Airflow Fan assembly is made of an aluminum housing unit, a 3-pin fan power plug for connecting to a header on your mainboard, and three 4cm fans. The DAF is fairly quiet, only being heard once the CPU and chipset fans were stopped. Corsair claims this assembly will lower module temperatures by roughly 4C, allowing for higher overclocks. But in our tests, the DAF did nothing to increase our overclocks.

The case used during our testing has a 250mm side fan, which would effectively cool the ram modules just as well with or without the assembly; so we're not surprised that installing it didn't net us any higher FSB speeds. The bling addict in us longs for clear bladed, LED illuminated fans but alas, it wasn't meant to be. About the only compatibility issue we can come up with is if you have a motherboard with DIMM slots running perpendicular to your graphics card slot. Those of us with newer motherboards shouldn't have this issue, as the manufacturers have been moving away from this practice for some time, now that longer video cards are coming out.

When first booting these modules in our EVGA 680i mobo, there was a brief moment of worry. Anyone keeping up with the 680i series of boards knows you’re just as likely to get a long beep indicating a memory error as you are a beep for booting into your OS. The Dominator modules booted right up with the EPP chip doing its job of setting the BIOS to the proper voltage, a problem with various motherboards, and the proper timings for stock performance. However, trying to boot without EPP enabled set our memory voltage to 1.8V, a voltage these modules flat out refuse to boot with. They will work at 1.9-2.0V, but as soon as we started cranking up the FSB or dialing down on the timings, we would get errors in Memtest within 60 seconds. After setting them to 2.1V, we were able to hit the stock FSB of 1066, but once again, memory errors and various boot issues were the norm. Once set to 2.2V, these modules simply loved doing their job and worked harder than Al Gore trying to convince us greenhouse gases are killing polar bears.

Overclocking Disclaimer

First and foremost, when purchasing a kit of RAM (or any component, for that matter) overclocking results are never guaranteed.  Many factors can influence what a kit is capable of, including complimentary components (motherboard, processor, power supply), user experience, and simple luck of the draw. Two identical systems will not necessarily yield identical results, and anything beyond specification should be treated as a bonus, not an expectation.

As we are about to turn the page to our performance benchmarks, let's take a look at how our tightest timings were set.

We tried tightening the timings on the ram at its stock speed of 1066, attempting to run 5-4-5-12. However, it  wouldn't pass Memtest with those timings. At stock settings, no matter if EPP was enabled or disabled, we simply could not tighten the timings without errors. Once we got beyond 1066 and reached our memory max of 1157, we could tighten the timings without issue.

In attempting to pummel this kit into the 1175 zone, we ran the voltage up to 2.4V. Upping the voltage beyond the recommended 2.2V did little for our efforts, and certainly not enough to risk damaging the modules. While set to 2.4V, the modules would benchmark at 1175, but fail Memtest. Random freezes during extended benchmarks, like SuperPi to 32m, quickly convinced us to lower the voltage and FSB to more reasonable levels.

We didn't reach 1175 like we wanted, but 1157FSB is far from lackluster. To illustrate, if you step up two price brackets ($80) to PC2-9136, which runs at an 1142FSB, you would still have higher performance at the PC2-8500CD5F's overclocked settings.

We can see from the table above that going from a 1T command rate to a 2T command rate nets mixed performance from DDR2-667 to DDR2-800. When DDR2 memory was first available, this 2T command rate was the cause of no small amount of complaining by the enthusiast community. The new DDR2 standard was slower than DDR due to higher timings and the 2T rate. Soon, DDR2-667 with tighter timings and a 1T command rate quelled much of the negativity surrounding the new format. Once we hit a 1066FSB, the pure horsepower from a higher FSB overrides the torquiness of the command rate allowing for a higher data transfer rate. Also, you can see that once the CPU is overclocked to its maximum, along with the memory overclock, the results for the benchmarks are vastly improved. In keeping the CPU at stock speeds for the first four columns of results, we can pinpoint the actual gains simply from changing memory FSB speeds

Our Recommendation

If I'd have to find a negative, it'd be the price. At well over $300 this kit is not for the faint of heart, and considering the rated speed, might be a tad overpriced. However, those looking for the best overclocking kit around should definitely add this kit to their list.

Other Reviews of Note

Its always nice to have more than one opinion on a component before you spend your hard earned money. For one, We may see something others missed, or vice versa. As with all reviews published at HardwareLogic, we'll not only give you our recommendation, but also point out some reviews from some other great sites around the web.

N/A