Articles: Mainboards

Bookmark and Share

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 ]


Considering the ECS NF650iSLIT-A mainboard didn’t do very well during CPU overclocking, didn’t show very good results during memory overclocking because of not very high maximum Vmem of only 1.95V and couldn’t really work at 1T Command Rate, it has very high chances of failing the comparison with any other mainboard. So, we decided not to perform any performance tests for that reason. There is only one question left that needs to be answered: how the graphics card working at half the nominal speed, as PCI Express x8, affects the performance of ECS NF650iSLIT-A mainboard?

To answer this question we took abit Fatal1ty FP-IN9 SLI. Since both mainboards are based on the same chipset we could emulate identical working conditions: overclocked the CPU to 375MHz FSB, set identical timings, graphics card working frequencies, 4x full-screen anti-aliasing mode and 16x anisotropic filtering mode set in the application properties. The only difference was the PCI Express speed: the graphics card on abit mainboard worked at PCI Express x16.

Frankly speaking, I didn’t expect to see any difference. In fact, you don’t really see it in the 3DMark06 total score and Shader Model 2.0 tests. However, in Shader Model 3.0 test and in F.E.A.R. XP game the performance differences are visible.

Although we have to admit that the difference is very small and may be consumed by the measuring error delta. If we have to really look for the performance difference in special benchmarks, then in real applications the users will hardly notice anything at all. So, we can consider it an insignificant drawback altogether.

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 ]


Comments currently: 2
Discussion started: 05/29/07 04:46:19 PM
Latest comment: 05/30/07 02:28:02 PM

View comments

Add your Comment