During our benchmarking sessions we used the conFiguretion listed below. Between our benchmarking sessions the only variable device was the motherboard being tested.
 
Hardware Configuration

• BIOS version: 54AW1P34, January 19th, 2006.
• Motherboard revision: N/A.
• Processor (performance benchmarking): Athlon 64 3800+ (2.4 GHz) with 512 KB L2 cache, NewCastle core.
• Processor (overclocking): Athlon 64 3500+ (2.2 GHz) with 512KB L2 cache, Venice core.
• Cooler: AMD In-a-Box.
• Memory: Two Corsair PC4000 TWINX1024-4000PRO modules with 512 MB each, installed on DDR Dual Channel configuration (3-4-4-8 1T timings).
• Hard Drive: Samsung SpinPoint SP0411N (7,200 rpm, 40 GB, ATA-133).
• Video Cards: Geforce 6600 GT 128 MB (Leadtek), Geforce 6200 128 MB (Leadtek)and GeForce 6200 TurboCache 64 MB 64-bit (XFX).
• Video resolution: 1024x768x32@75Hz.
• Power Supply: Seventeam ST420BKV-03F.

Software Configuration

• Windows XP Professional installed using NTFS
• Service Pack 2
• DirectX 9.0c

Driver Versions

• nVidia video driver version : 84.21 WHQL.
• nVidia nForce driver version: 6.86 WHQL (nForce 4 Ultra).
• nVidia nForce driver version: 8.25 WHQL (nForce 6100 + NF410).
• ATI driver version: Catalyst 6.6.
• Audio driver version: Realtek 3.89.

Used Software

• SYSmark2004 – Patch 2
• PCMark04 Business 1.3.0
• 3DMark2001 SE 3.3.0
• 3DMark03 Pro 3.6.0
• 3DMark05 Pro 1.2.0
• Doom 3
• Quake III Arena 1.32

We adopted a 3% error margin. Thus differences below 3% cannot be considered relevant. In other words, products with a performance difference below 3% should be considered as having similar performance.