Thread: Typically generations of tested motherboards

How We Tested Testing! That must mean lots and lots of statistics and numbers - everything to do with performance. It's the case with this motherboard test, but it's not only about the numbers either. It's a little known fact to many that performance in motherboards for all desktop applications and even gaming for that matter doesn't really differ a whole lot. Typically from past generations of tested motherboards, a difference of one to four per cent is expected in applications and games alike, as long as there is a common processor or a graphics card on the system. In our case, we have motherboards with onboard graphics rendering capabilities and there is bound to be a difference of more than just one to four per cent. TIle test process although pretty similar to previous times, had a little more weight¬ing assigned to the game and graphics intensive benchmarks.

We ran two fairly old games to get our scores. It wouldn't make sense to attempt running extremely GPU intensive games such as Crysis on the onboard graphics cards. We ran Doom 3 and Company of Heroes at resolutions of 800 x 600, 1024 x 768 and 1280 x 1024. These games would give a substantial frame rate to measure, and users who do not plan on purchasing graphics cards would be running such games at lower resolutions on their 17-inch screens. Other than the game benchmarks, our usual set of regular benchmarks were used to measure perfor¬mance of applications on the desktop - PC Mark 05 and the synthetic benchmark SiSoft Sandra, were also used. Weightings were given to the tests and we decided to set lower priority to SiSoft Sandra, the synthetic benchmark which has no direct rela¬tion to real world performance most of the time, but can be used to measure and compare statistics between motherboards of each platform.

Most of the other weighting was given for every¬thing other than performance. Motherboards had to have plenty of features - be it loads of SATA and USB ports or support for multiple graphics cards. Because these are motherboards with on board graphics solutions, it's OK to assume that most of these motherboards come with good HD video decoding capabilities.



Throughout the evaluation, we looked at the design and layout of the boards. We were looking for clean spacious layouts that allowed easy instal¬lation of the processor and the graphic card with no SATA ports located under the slot. Solid-state capacitors were also a factor, which would typi¬cally mean a better chance of a long living board. Graphics cards get hot and people know that. Now imagine something like that on your motherboard. Some of boards came with great coolers and heat dissipation units such as heatsinks and heatpipes to help keep the overactive chipset temperatures under control. Everybody loves freebies, so we also gave points to motherboards that came with additional fans or coolers that would enhance the performance or stability of the boards in any way. Freebies also included the fun stuff - games and applications to use on the system.

As for the test rig, we chose the fastest pro¬cessors for either platform that we could find. We used an AMD Phenom X4 9950 for the AMD motherboards and an Intel QX9650 for the Intel ones. For the storage drives, we used two identical Seagate 7200.11 500 GB. These drives are very fast, affordable and would probably be the kind of drive someone buying a budget system with an inte¬grated graphics solution would use. With the same idea of affordable hardware in mind, we stuck in 2 GB DDR2of Corsair Dominator RAM on the boards. Two Corsair HX620's, which we admit are a little too powerful, for the setups were also used.


It must also be clarified that the processors from AMD and Intel are not identical in performance, so it would be wrong to compare all the scores of the Intel boards with the AMD ones. Although the pro¬cessor and motherboard were the only hardware differing, the two tests were handled separately.