nForce 2 and KT400 in a duel: Asus A7N8X Deluxe against EPoX 8RDA + and EP-8K9A2 +

nForce 2 and KT400 in a duel: Asus A7N8X Deluxe against EPoX 8RDA + and EP-8K9A2 +


After we had already examined the EPoX 8RDA + in an extra article at a wide variety of clock rates In order to check on the one hand the overclocking capabilities of the board and on the other hand the compatibility to a possibly appearing Athlon XP 'Barton' with FSB400, our two other candidates had to undergo this test. The nForce 2 seems to have a decisive advantage in purely theoretical terms, since the AGP and PCI clock can be set to 66MHz and 33MHz respectively - a feature that is otherwise only found on Intel boards.

Beyond that, however, our article threw another oneAnother important finding about the nForce 2: The chipset can only develop its true power with a memory clock running synchronously with the FSB. Clock ratios like 266: 333 simply make no sense. More details can be found in the article .


As briefly mentioned in the introduction, the EPoX 8RDA + circumvented the FSB400 hurdle without the slightest navigation error. The board ran stably and extremely quickly through the entire benchmark course. And that despite passive chipset cooling, which doesn't even cover the entire heat spreader of the nForce 2. Respect!

EPoX 8K9A2 +

We were particularly surprised by the 8K9A2 + from EPoX, because we would have thought much less of the KT400 chipset used. But the board ran with an FSB of 200 MHz (FSB400) and two bars Corsair PC3000CL2 with DDR400 just as problem-free as the 8RDA + before. Remarkable when you consider that VIA has not even validated DDR400 with a module. However, we do not want to generalize this service. You should also consider the fact that the board was not exposed to any major thermal loads when it was lying freely on the floor. Since the mainboard does not have a 1: 6 divider, the PCI clock at 200MHz FSB ran at 40MHz and the AGP bus at 80MHz, which should give the mainboard a small speed advantage compared to its nForce 2 colleagues.

8K9A2 + FSB400

Asus A7N8X

In view of the performance of the 8RDA +, we were extremely relaxed about the FSB400 test on the A7N8X. 'Same chipset, same result' was our motto. Yes, far from it! With the Bios 1001E used for benching, the board wanted to start at 10x200MHz, but was noticed unpleasantly just a few seconds after booting.With every mouse click, the image on the monitor disappeared for a fraction of a second and when we tried to control the WCPUID settings, the display finally got out of control. And that although we had locked the AGP clock to 66MHz via BIOS. And even with the setting Auto or 50 or 67MHz, the same errors appeared and the 3DMark as well as every other, computationally intensive application said goodbye to the screen surface after a few seconds and hundreds of image errors.

Asus A7N8X FSB400 malfunction

During the test A new BIOS version appeared with the 1001G, which was flashed into the EEPROM immediately. However, this update did not bring us any improvement either. In addition to the known bugs, the board now booted from time to time with the setting of 10x200MHz with 7x200MHz and finally drove the tester crazy. In the end we were satisfied with smaller rolls and found the stability limit at 191MHz FSB. We want to rule out a malfunction of the 'AGP-Fix' and come to the conclusion that our model did not want to participate in a 200MHz FSB. Consultations with colleagues on other sites yielded results of 185-200MHz, which should make it clear once again that the result of a single chip does not have to become the law of the series.

On the next page: Benchmarks