OCZ PC-3500 Platinum EB Dual Channel Memory Kit

Testing Methodology

To test each kit I ran three different benchmarking programs at a number of different speeds. The first benchmark was AIDA32 3.9. It does a basic read and write test to gauge memory bandwidth. The second test I ran was SiSoftware's Sandra 2004. I ran the Combined Performance Index Wizard through to completion, although I only used the memory score in the chart. The third benchmark was Futuremark's PCMark04 Pro. I ran the full test suite to completion, all 44 tests, before taking just the memory score for use in the chart.

If the RAM could not finish a test at a given FSB using the SPD setting in the BIOS, I stopped testing it then and there. Why didn't I tweak the timings? Because if you loosen the timings on low latency RAM you end up with high latency RAM, which is not how the product is marketed. It is my feeling that the reason someone buys low latency RAM is for the low latency aspect of the RAM. If he or she had wanted to buy faster RAM with higher latency, that's what they would have bought in the first place.

Testing Setup

Here's the system I used to do the testing of each kit. Only the FSB was raised during testing, the other clocks were kept at their base levels, including the video card:

CPU: Intel Pentium 4 2.6C
Motherboard: Albatron PX865PE Pro II
Motherboard BIOS: 1.10
Heatsink: Scythe Kamakaze
Fan: Thermaltake Smart Fan II
Video Card: ATi Radeon 9700 Pro 128MB
Video Driver: Catalyst 4.6
Hard Drive: WD 80GB ATA 100 7200RPM Caviar Special Edition
Power Supply: Seasonic SS-400FS

If you have any comments or questions regarding this review, please post them in this thread on our forums.

Select a Page Page 1: Introduction Page 2: Testing Methodology Page 3: AIDA32 Test Results Page 4: SiSoftware Sandra 2004 & PCMark04 Pro Page 5: Conclusion