We originally did most of our testing with the BIOS that shipped with the board, and performance was a lot lower than we expected, given the results we had on-hand from the
MSI KT4 Ultra. In fact, these initial results were lower than both the KT4 Ultra numbers, and the
KT3 Ultra 2 (KT333) numbers that we used as a reference. Midway through our testing, Soyo released a BETA BIOS that they urged reviewers to try out: this BIOS was released in response to reviewer concerns over DDR400 performance, and our tests found that the new BIOS gave better performance when using DDR400.
Test System Specs:
AMD 2100XP
512MB Corsair XMS3200 CL2 RAM (CL2 at DDR333)
512MB Corsair XMS3500 CL2 RAM (CL2 at DDR400)
The KT4 Ultra system which we are comparing used the same components.
Unless otherwise noted, tests were run using the XMS3200 CL2 RAM at 333Mhz CL2 setting.
Sisoft Sandra
   |
| Sandra CPU, Memory, & Multimedia Marks |
All of the Sandra results show the Soyo KT400 board slightly behind the KT4 Ultra. These are such slight differences that you shouldnt base too much of your conclusion on this data the biggest difference is 2.7% in the RAM Float benchmark. Whats important to note from these charts is that the Soyo KT400 uses the CPU as efficiently as the MSI KT4 Ultra. Its the RAM thats underperforming, roughly 1-3%. This point will come across more strongly in the PCMark test scores.
SysMark 2002
Looking at these results, you can see that the KT4 Ultra has a very slight lead in the ICC tests, but in the Office Productivity test, which simulates general office applications use, the Soyo Dragon Ultra falls behind by 18 points, which is a 12.5% difference!
PCMark:
The PCMark results show some of the most dramatic performance differences between the two boards. As seen with the Sandra results, the CPU scores are very, very close, but look at the sizable gap between the Memory scores: thats a near 7.5% difference!
forums
send article 
hardware newsletter 
article comments (2)
Digg this!
del.icio.us