Review: EpoX 4PDA2+ i865PE Springdale

System setup and notes

• ATi Radeon 9700 Pro (324/310)
• 2 x 256MB Corsair XMS3500C2, run at 2-6-2-2 @ DDR-400 for the 4PDA2+ and SN45G in DC modes and a single 256MB stick for the KT600 in SC mode
• Liteon 16x DVD
• Samcheer 420w PSU
• Samsung 181T TFT monitor
• Akasa Silver Mountain cooler
• Intel reference cooler

• Windows XP Professional Build 2600.xpclient.010817-1148
• DirectX9.0a
• Intel 5.00.1012 chipset drivers
• Hyperion 4.46 VIA service pack
• NVIDIA nForce 2.03 drivers
• ATI CATALYST 3.4 drivers and control panel (6343s)
• Pifast v41 to 10m places
• Lame v3.92 MP3 encoding with Razor-Lame 1.15 front-end using U2's Pop album
• SiSoft SANDRA 2003 (9.44 release with SP1)
• Hexus SETI benchmark
• 3DMark 2001SE
• UT2003 Demo (Build 2206)
• Comanche 4 benchmark
• Serious Sam 2 Demo
• Quake 3 v1.30 HQ

Notes

Two Intel and two AMD chipsets will fight it out to be crowned the performance champion of our select little group. The 4PDA2+ would sometimes take a while to boot from a cold start. That problem, however, seemed to be alleviated somewhat with a flash to the 6/18 BIOS. As mentioned before, the bundle shipped without a driver CD but obtaining the drivers via another PC connected to the net proved to be as simple as a Google search.

Stability was good. The combination of a stable OS and mature chipset manufacturing process should usually allow for top-notch stability. The board had no problems in running our standard 2-6-2-2 memory timings. Adding a further two Corsair XMS3500 C2 modules resulted in a slight relaxation to 2-7-2-3 (tRAS=3).

Once again, explicitly stating clock speeds should inform us whether certain manufacturers are gaining an artificial benchmark lead.

3222.6MHz / 201.4FSB - EPoX 4PDA2+

3208.1MHz / 200.5FSB - DFI LANPARTY PRO875

2205.21MHz / 200.47FSB - Shuttle SN45G nForce2 (AMD)

2205.2 / 200.47FSB - VIA KT600 (AMD)

The EPoX steals around 15MHz on the DFI Canterwood board. ~ 0.5% clock speed difference is enough to make two boards based on one chipset appear statistically different. We're benchmarking four different chipsets so that is less of an issue. We'd wish that manufacturers would have some form of FSB consistency. A quick note about the comparison KT600. Our sample featured an initial BIOS that wasn't geared towards pure performance. We expect its numbers to improve with subsequent BIOS releases.

Lastly, a switch to a 3.2GHz CPU and a Radeon 9800 Pro's death means that previous boards' results cannot be compared to the results obtained here. An ATi Radeon 9700 Pro takes up position in the AGP slot.

FSB overclocking went better than expected. We normally expect at least 250FSB from each and every i865 / i875 board. Anything above this can be considered a bonus. The 4PDA2+ managed our basic 250FSB with ease. 275FSB came and went.

290+ FSB was semi-flakey but we can't complain at 288FSB with utter stability. Well above average, we'd say.