EPoX 8K7A+ Evaluation

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Game benchmarks

The game benchmarks have been run in a number of different resolutions. Remember this, the 640 x 480 x 16-bit tests are the ones that will test the performance of the CPU, memory and motherboard, if it makes any difference. The settings used can be found here. However nobody plays at 640x480x16-bit if they don’t have to, so I’ve included the higher resolutions in order to get a feel for what you will achieve in actual play. The high resolutions are only going to test how good the video card is, and are just game play indicators for your reference. The scores to concentrate on are the low-resolution scores.

First up we have the venerable Quake 3 Arena. This is an early version using an early demo. This test, and the next few will show you how the 8K7A+ performs against the competition in older “legacy” (for want of a better word) games.

The EPoX 8K7A squeaks home, followed by the KT7A, with the KA266 trailing further behind. But even at the Normal settings, 175 frames per second is nothing to be ashamed of. Notice also, that even at the “Normal” setting, the game is still being limited by the chipset/memory or CPU. The relative positions are unchanged.

The High Quality setting is what we mean by by a video card limitation. See how the scores are equal. From the above graph, we know that the 8K7A+ is fastest and the KA266 is slowest. But at a resolution that people will realistically play at, it doesn’t matter which motherboard you choose because the video card is the bottleneck.

Next up we have Expendable. In contrast to Q3A, which is an OpenGL game, Expendable is a DirectX game. Does this make a difference?

Nope, the order is the same, 8K7A+ out in front, with the KA266 and KT7A trailing. But Expendable has a neat little trick – it can tell us the minimum framerate scored as well. This is a more important figure in game play, as it will tell us if the game will stutter just when we least need it.

Again the 8K7A+ is fastest, with very little between the KT7A and the KA266. It is interesting that at the high resolutions, this game isn’t quite video card limited like Q3A is. And in this case, the scores are in order of latency, best to worst , and the KT7A wins.

Also using an older game engine is MDK2

Another First Person Shooter, and it follows a similar trend to Q3A. At higher resolutions, the KT7A wins out, even over the EPoX, but as soon as we bump the resolution – things even out. The fact that the KT7A wins even at the “Normal” settings tends to suggest that MDK2 is latency sensitive, whilst being fairly indifferent to bandwidth. This is further supported by the scores from the KA266 which we know has much worse latency than either the KT7A or the 8K7A+, while the KT7A and 8K7A are fairly close together in terms of latency (as are their scores). Notice also, that there is virtually no difference in scores from the “Fastest” setting to the “Normal” setting, which suggests that this game as actually CPU limited at low resolutions.


Pages: « Prev   1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16   Next »

Be the first to discuss this article!