Menu
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 +

Benchmarks

SiSoft Sandra 2003 Pro - FSB400
  • Int. Bufferd:
    • EPoX 8RDA + FSB400/DDR400
      3.071
    • EPoX 8K9A2 + FSB400/DDR400
      2.927
  • float. Bufferd:
    • EPoX 8RDA + FSB400/DDR400
      2.870
    • EPoX 8K9A2 + FSB400/DDR400
      2,698

The lead of the nForce 2 over the KT400 at FSB333/DDR333 is 2.5%. Nowit is 5% and 6% respectively. The 8RDA + can therefore draw significantly more power from the higher clock rate.

Cachemem - FSB400
    • EPoX 8RDA + FSB400/DDR400
      215
    • EPoX 8K9A2 + FSB400/DDR400
      282

Cachemem also confirms that the nForce 2 behaves better with the FSB400/DDR400 despite the identical memory timing. While the nForce 2 was able to correct the latencies down slightly, the KT400 recorded a slight increase. A performance difference of 23.4% becomes 31%.

3DMark 2001SE - FSB400
Unit: points
    • EPoX 8RDA + FSB400/DDR400
      12,792
    • EPoX 8K9A2 + FSB400 /DDR400
      12.389

Both chipsets make good progress in 3DMark 2001. The KT400 with FSB400 and XP2400 + achieves the result of the nForce 2 with XP2400 + and FSB333. However, this is now 3.2% in the front. With 333/333 it was still 2.9%. The result of the 8RDA + corresponds to that of the nForce 2 with FSB333 and XP2600 +.

Quake3Arena - FSB400
  • 1280x1024 Max details:
    • EPoX 8K9A2 + FSB400/DDR400
      199.70
    • EPoX 8RDA + FSB400/DDR400
      196.50

The KT400 in front? After an initial phase of uncertainty, however, we remembered the cause of this phenomenon. The 8K9A2 + does not have a fixed AGP clock and thus ran with 80MHz AGP (standard: 66MHz) in this test. Quake3Arena seems to benefit particularly from this and enables the KT400 to do thisonly victory in our benchmark course.

Comanche 4 - FSB400
    • EPoX 8RDA + FSB400/DDR400
      49.15
    • EPoX 8K9A2 + FSB400/DDR400
      46,31

In Comanche, too, the nForce 2 with FSB400/DDR400 can further expand its lead from 5.7% to 6.1%. Such minor differences are of no consequence in everyday practice. However, they illustrate the good scaling of the nForce 2 even at higher clock rates.

Unreal Tournament 2003 - FSB400
  • FlyBy:
    • EPoX 8RDA + FSB400/DDR400
      161,94
    • EPoX 8K9A2 + FSB400/DDR400
      160,68
  • Botmatch:
    • EPoX 8RDA + FSB400/DDR400
      67,51
    • EPoX 8K9A2 + FSB400/DDR400
      63,56

Both also scale in the UT2003 botmatch Board well with the increased clock rate and almost achieved the results with an XP2600 + (FSB333). In the graphics card limited FlyBy the results are even higher.

WinACE 2.11 - FSB400
Unit: minutes, seconds
    • EPoX 8RDA + FSB400/DDR400
      3:15
    • EPoX 8K9A2 + FSB400/DDR400
      3:32

In WinACE, the KT400 catches up with the nForce 2 for the first time and can keep track of the deficit from earlier ( FSB333/DDR333) from almost 10% to 7%.

Overall, amazing results. To theone that both boards survived these high clock rates without any problems. On the other hand, that they both scale to a greater or lesser extent. The KT400 roughly achieves the results of the nForce 2 with FSB333/DDR333 with the same CPU clock and thus proves that the poor performance with DDR400 with synchronous control is blown away. Disappointing that the A7N8X did not pass this test and that we have to refrain from making a general statement about the FSB400 suitability of the nForce 2.

On the next page: Conclusion