AMD Athlon 64 FX-57

System Setup

ASUS A8N-SLI Deluxe Motherboard (NVIDIA NForce 4 SLI chipset); XFX GeForce 6800 GT PCI-Express (operating at its default clock speeds of 350/1000MHz); 2x 512MB Corsair TwinX10224-3200XL @ 400MHz, 2.0-2-2-10 1T; Western Digital Caviar WD2000, 200GB SATA Hard Disk Drive; LG40x DVD-ROM drive; OCZ Powerstream 520W PSU; Windows XP Professional Service Pack 2; NVIDIA NForce 4 Standalone platform drivers version 6.53; NVIDIA Forceware 71.89 WHQL.

Processors:

• AMD Athlon 64 FX-57
  
• AMD Athlon 64 X2 4800+
  
• AMD Athlon 64 FX-55
  
• AMD Athlon 64 4000+

Game Performance:

• Doom 3: demo1.dem; 640x480 Low Quality; 1600x1200 0xAA 8xAF High Quality.
  
• Far Cry v1.31: Fort demo; 800x600 Low Quality; 1280x1024 4xAA 8xAF Maximum Quality.
  
• Unreal Tournament 2004: custom Ons-Dria demo; 800x600 Low Quality; 1600x1200 Maximum Quality.

General Performance:

• DVD Encoding: DVD Shrink; default settings, 50.0% shrink of The Naked Gun DVD.
  
• MP3 Encoding: RazorLAME v3.97; encode Moby - Play album from .WAV to 192kbps .MP3.
  
• File Compression: WinRAR; compress UT2004 root directory.
  
• Scientific Calculation: SuperPI 2M calculation.
  
• Memory Bandwidth: SiSoft Sandra Lite 2005; Unbuffered Memory benchmark.

Game Performance

Summary:

In general, the FX-57 delivers no surprises in our gaming benchmarks. At the lower quality settings, we see that the Athlon 64 FX-57 delivers an incremental performance increase over the Athlon 64 FX-55 and 4000+. This is with the exception of Unreal Tournament 2004, which shows a greater performance increase between FX-55 and FX-57 than the one we experienced between the FX-55 and 4000+.

However, when the image quality was increased to more realistic levels, we found that both Doom 3 and Far Cry did not show any great performance difference between the CPUs, while Unreal Tournament 2004 showed a small performance increase as CPU speed increased. Unreal Tournament is generally very CPU limited, even in the most graphic-intensive sections of the title.