Test results - bandwidth and latency

First in our benchmark suite are the bandwidth tests 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 nice benchmark suite, in which you can compare your system to a range of other reference systems. The Bandwidth is one of the available benchmarks. Alternatively, for bandwidth measurements, we can use the Everest application from Lavalys, which consistently gives higher results, but the scaling is (almost) the same as Sisoftware's Sandra :



A nice and clear line upwards, as you can see : logic is respected, with higher speed memory giving higher bandwidth, and extra front side bus speed providing a healthy boost. Striking here is the extra bandwidth you get at 500FSB, going from 1600Mhz to 2Ghz on the memory.

Next up is the Latency result, for which we use Everest once more:



Not entirely a surprise, but still. We know that latency is sacrificed in favour of bandwidth, but here you clearly see the latency at 1333Mhz (333FSB x 9) equal or better than the latency at higher FSB (for the same cpu speeds). 500FSB does generate much better results, but remember that the top two results cannot be compared to the bottom four.


Test results - synthetic benchmarks

As we still test on Windows XP (SP2), we can't use the Vantage benchmark. There is of course the older PC-Mark 2005 which gives a nice indication of overall system speed.



This results might surprise some, but they have been repeatedly checked. The result actually confirms the latency readings, showing that the basic front side bus test, with low latency ram, is actually the fastest way to go. This is in line with the findings of Tony Leach, which he details in the OCZ support forums. Remember, we're primarily testing the motherboard here, with the ram on the second row.

Next up is 3D Mark 2006 :



3D-Mark 2006 proved to be a benchmark giving no improvements at all at different memory speeds, and our stock cpu speed. At 500FSB there are small gains to be had by overclocking the memory.

Let's follow the 3D Mark with SuperPi, very well known to follow bandwidth and latency results rather closely:



Superpi likes bandwidth, you might remember that. The superpi scores follow the bandwidth improvements we see in the first chart of this page. The 1M calculation is simply too short to give any real differences.


Test results - real-world application benchmarks

Let's first look at the Cinebench results, which mimic the Cinema 4D application. We're up to version 10 now, have a look :



Differences are very small, but we get the same results as our PC-Mark 2005 benchmark. The Low latency score is the best (at 3Ghz), proving that this is the speed you want to run your PC at if you own a P45 DDR3 motherboard.


Let's move on to the H.264 test. H.264 has created a lot of momentum after the inclusion in Apple's Quicktime, and is considered an important video standard now. Encoding/decoding H.264 is a very cpu-intensive task, and we'll be measuring the impact of the memory subsystem on H.264 performance next:



Very, very small differences here, I actually reran this benchmark several times and the shown result is an average. Sometimes one was faster, sometimes the other. At 500FSB the 2Ghz setting did outperform the 1600Mhz setting consequently.

Finally, a bit of Crysis to get a feel for gaming performance:



Again, small differences only. We used a 7300GT card, not exactly a gaming card, but used it to show you the impact of cpu and memory. We tested the cpu test at 1024x768 setting, at low quality and DX9 (Win XP). The impact of memory and cpu is low as you can see, with barely a difference for a cpu at 4Ghz in relation to the 3Ghz settings. Crysis is gpu-bound, and you don't need the fastest memory, you need the fastest gpu.

On to the final words and conclusions ->