System setup and notes
Here's a quick rundown of the test system should you wish to compare benchmark results with your own.- AMD Barton XP3200+ S462 CPU (2200MHz / 200FSB)
- Intel Pentium 4 3.2GHz 800FSB CPU (3.2GHz / 200FSB)
- EPoX 8RDA3G nForce2 Ultra 400 (21/7/03 BIOS)
- EPoX 4PDA2+ i865PE Springdale
- MSI KT6 Delta-FIS2R KT600
Other components
- ATi Radeon 9
Software
- Windows XP Professional Build 2600.xpclient.010817-1148
- DirectX9
Notes
Normally this section just deals with the methods by which benchmarks will be conducted. Sample motherboards usually work just fine. However, with this sample, a few idiosyncrasies presented themselves. Firstly, several Barton and Athlon XP Thoroughbred chips were tested. None of these managed to boot the board at 200FSB; the running speed of the nForce2 Ultra 400 chipset. The highest stable frequency was 198MHz FSB. Raising the chipset voltage did nothing to solve the problem. After much hair-tearing, it emerged that the board had been shipped with the JCLK (clock select) jumper in the wrong position. Moving it over a notch rectified the problem. This may just be an isolated incident but we feel that it's worthy of attention here.
Another aspect centred around relatively poor voltage regulation, if the BIOS was to be believed. Setting a Vcore of 1.65v resulted in a Windows load voltage of between 1.55 - 1.58v. EPoX's previous boards have been pretty good on the voltage lines in the past. Even DDR voltage wasn't immune to under-volting. Setting 2.90v DDR in BIOS gave around 2.77v under load. It's difficult to say whether this is a board specific or widespread issue. We're just thankful that EPoX gives the user enough voltage to play around with in the first place.
Lastly, and perhaps most curious of all was, on occasion, the board's ability to manually change the FSB in Windows. Setting it at, say, 2200MHz, it would sometimes update the speed on the fly from anywhere between 1600MHz to 2200MHz, according to the latest build of CPU-Z. This was only discovered after inconsistent benchmark runs. Yet if the board was rebooted, it maintained a rock-solid frequency. The benchmarks that follow are based on a stable frequency.
After most of the small niggles were ironed out, the sample 8RDA3G was extremely stable at 200MHz FSB+. FSB testing, with a Vdd set to 1.8v (1.76v under load), seemed to suggest that it was stable right up to 226FSB. That's par for the course these days, we feel. The board ran a standard XP3200+ CPU at 2205.1MHz; 5.1MHz above official specification.
