Test methodology

For every benchmark or application two tests were performed, at two different speeds. Reason for this is the speed rating of our Dominator test kit, which is PC8888 or 1111Mhz DDR. The memory speeds which can be applied on our motherboard are 533Mhz (1:1), 667Mhz (4:5), 800Mhz (2:3), 889Mhz (3:5) and 1066Mhz (1:2). All these settings can be applied from the standard FSB (front side bus) of 266Mhz. As you can see, 1111Mhz is not listed.

In a first test we want to run the memory at its rated speed. To be able to do this, we have to overclock the CPU slightly. In this case an overclock was chosen using the 8x multiplier of the CPU. The front side bus was brought up to 370FSB which results in a CPU speed of 2.96Ghz, up from the standard 2.66Ghz. When you further select the "800Mhz" or 2:3 memory ratio we reach the rated memory speed of 1111 Mhz (give or take 1Mhz).

We could also have used a higher multiplier (9 or 10) and ram ratio (3:5 or 1:2), and this would have meant a (far) lower bandwidth. For the "cheaper" memory kit I cannot use the same 370Mhz FSB setting, as the PC4300 simply would not run. Instead, to make things comparable, I used the maximum multiplier of 10 for this ram and ran it at 296FSB (x10). Most PC4300 ram should be able to run at the slightly overclocked setting of 296Mhz. One might argue that this test benefits the Dominator ram more than necessary, but I call this the prerogative of a high end memory kit: being able to run higher front side bus as well as higher memory speeds.

The second test was performed at the standard speeds of 2.66Ghz, and with standard selectable memory settings of 533Mhz (1:1) and 1066Mhz (1:2). This test is completely comparable for both sets (same speed and front side bus), but in this case the Dominator ram is at a disadvantage as it is running below specifications.


Bandwidth and Latency

To measure the bandwidth and latency performance of the memory kit we use two standard applications.

First is the bandwidth for which we use SiSoftware's Sandra application. The System ANalyser, Diagnostic and Reporting Assistant is an information and diagnostic utility which has a very handy benchmark suite, in which you can compare your system to a range of other reference systems. The Bandwidth is one of the available benchmarks :



It doesn't come as a surprise of course that the Dominator pulls away with a 36% better bandwidth than the PC4300 kit at the 2.96Ghz setting. Even at 2.66Ghz, when the memory does not run at its full capacity, the difference is still around 20%. One additional thing to bare in mind though is that the PC4300 kit cannot be brought to more bandwidth by increasing the front side bus, whereas the Dominator could theoretically offer even more bandwith when running at a lower ram ratio (1:1 for example) and a higher front side bus, providing that your motherboard can follow suite of course.

Second up is the latency test for which we use the Everest application from Lavalys.



Latencies are much less depending on the front side bus than bandwidth, and the Dominator clearly takes the lead here. Latencies are not the strongest point of an Intel platform, but the sheer speed of the ram makes up for much in this case.

Application performance comparison

As we mentioned on the first page, we will not go very deep into the performance aspect: full data numbers will be presented in our upcoming high end memory roundup. In the roundup the Dominator will be compared with some other high end memory kits. However, we do not want to keep you hungry too long, so we'll do things a little different this time.

In the following graph you can see the performance increase that can be expected when you upgrade your memory from PC4300 @ 4-4-4-15 settings to the Dominator PC8888C4 kit.



So far for the results at a CPU speed of 2.66Ghz, let's look at the rated speed of the Dominator next at 2.96Ghz :



I'll leave it up to you if you like these numbers or not... Performance increases can be expected from only 1 up to almost 10%, depending on speeds and application. I'd say you would be able to expect an average increase in performance of about 5%. But once again, when interpreting these numbers, bare in mind the following: if you want to overclock your PC, you'll need to find a way to increase your front side bus and unless you have an extreme edition processor (which has an unlocked multiplier), you'll need a lot more than PC4300 as well.


Before we wrap it all up, let's have a look at the overclocking capacity of the Dominator on the next page ->