The Platforms con't.
Next the AMD based Socket 939 chipset MSI K8N Neo4 Ultra.
- Motherboard: MSI K8N Neo4 Ultra
- Processor: AMD 3500+ Winchester
- CPU Cooling: Thermalright XP90 w/Panaflo 90mm fan
- Memory: Mushkin Level 2, Version 2 (TCCD) 2X512 DDR
- Graphic Card: eVGA 6600GT (PCI-E)
- Hard Drive: Western Digital IDE 120GB 8MB cache
- PSU: Antec True Power 550watt
- Windows XP PRO
- Direct X 9.0C
- nVidia 71.84 drivers
AMD's Socket 939
Methodology
We'll start with default CPU and memory speeds for both platforms. DDR timings @ 2-2-2-5 for the Intel platform and 2-2-2-7 for the AMD platform. This is not meant to be an Intel or AMD bash review but for the purpose of easy identification during the benchmarks we'll identify the systems as Intel or AMD.
We'll then over-clock both systems to maximum stable performance and repeat our suite of tests. It's pretty straight forward finding the sweet spot on the Intel system. Either your CPU or DDR will be the limiting factor and in this case the 3.2C maxed out just over 3800 MHz and the DDR was able to match that speed at good timings of 2-3-3-6 and we were ready to go. There's a bit more to consider with the AMD system. Not only do you need to find your system's Max CPU and DDR clocks, CPU multiplier and managing your HTT around the 1000 mark also come into play. Max Vdimm being an issue best performance was CPU @ 2661 MHz and DDR at 2-3-3-7.
- Intel @ 238 FSB
- CPU @ 3808 MHz
- DDR 476 @ 2-3-3-6
Intel
- AMD @ 242FSB/ 11X Multiplier/4X HTT
- CPU @ 2661 MHz
- DDR 484 @ 2-3-3-7
AMD
Tests
CPU Performance: Super_PI 32M, hexus_pifast
DDR Performance: SiSoftware Sandra Unbuffered Memory Bandwidth Benchmark, Everest Home Read/Write/Latency.
Game Performance: Half Life2 (HardwareOC d13c17), Doom3 (Benchemall), AquaMark and 3DMark 2005. Both video cards ran at their stock settings throughout the tests. Antialiasing and anisotropic filtering were set to application controlled, vertical sync was disabled, and image quality was set to "High Quality" in the nVidia control panel.
