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+ (11x200)
- Intel Pentium 4 3.00GHz HT S478 Northwood CPU (800FSB)
- Shuttle AN35N Ultra 400 (AN35S930 BIOS)
- EPoX 8RDA3+ nForce2 Ultra 400 motherboard (29/04/03 BIOS)
- Asus P4C800 Deluxe Canterwood
- ABIT BH7 i845PE at 200FSB
Common components
- ATi Radeon 9800 Pro (380/340)
- 2 x 256MB Corsair XMS3500C2 run at 2-6-2-2 at DDR-400 for AN35N, 8RDA3+, P4C800, BH7, DDR-333 for NF7-S
- 41.5GB IBM 120GXP Hard Drive
- Liteon 16x DVD
- Samcheer 420w PSU
- Samsung 181T TFT monitor
- Akasa Silver Mountain cooler
- Cooler Master Fujiyama heatpipe cooler
Software
- Windows XP Professional Build 2600.xpclient.010817-1148
- DirectX9
- Intel 5.00.1012 chipset drivers
- NVIDIA nForce 2.03 drivers
- ATI CATALYST 3.2 drivers and control panel (6307s)
- Pifast v41 to 10m places
- Lame v3.91 MP3 encoding with Razor-Lame 1.15 front-end using U2's Pop album
- SiSoft SANDRA 2003 (9.43 release)
- Hexus SETI benchmark
- 3DMark 2001SE v330
- UT2003 Demo (Build 2206)
- Comanche 4 benchmark
- Serious Sam 2 Demo
- Quake 3 v1.30 HQ
Notes and issues
Here's where it gets interesting. We tried the AN35N with 2 pre-tested sticks of Corsair's excellent XMS3500 RAM. Not only have these sticks run as a pair at DDR-400 with 2-6-2-2 timings, they have done so with default volts. This has been tested on a number of cross-platform motherboards from various manufacturers. Imagine our surprise, then, when the board failed to load a fresh copy of Windows XP Professional with these timings (we always install the OS with relatively relaxed timings). Raising the voltage to the maximum 2.7v didn't cure our stability problems. Using another pre-tested pair of Mushkin EMS PC3500 modules didn't help either. Using any number of high-performance PC3200+ modules resulted in almost instant crashing when loading the OS. However, relaxing the memory timings to 2-3-3-7 cured all ailments. It appears as if the board is applying some odd latencies when set to 2-2-2-6 in BIOS. To confuse matters a little more, running at 166FSB caused no problems whatsoever, even with the tightest latencies. The sample board also had a problem restarting from 200FSB upwards. It was even money whether it would reboot correctly or not. In its defence, it did power-up perfectly each and every time.
We tried e-mailing Shuttle through their web-based form. We simply could have gone through our usual contact and had the board replaced. However, we wanted to see how the average consumer was dealt with. Unfortunately, our odd problem received no reply within the stated 24-hour period, or 72 hours for that matter. Further, there were no AN35N BIOS files on any of Shuttle's 'sites. We can appreciate that the latter might be the case due to the recent introduction of the board.
Lastly, we've kept the older nForce2 2.03 and ATi 9800 Pro 6307 drivers for parity's sake. The comparison EPoX 8RDA3+ was benchmarked with that set. We'd have re-benchmarked the EPoX with the newer versions of both drivers had it not unexpextedly died soon after official benchmarking and stress-testing had finished. We don't expect the Shuttle to benchmark as well as the EPoX, simply because we're not in the position to use performance-enhancing timings at 200FSB.
From an overclocking point of view, we managed to hit 216FSB with decent stability. Anything higher would cause the board to reset when placed under scrutiny.
