Overclocking and Optimizing:
After a few benchmarks and a 30 minute session of OCCT were executed at default settings I discovered something worth addressing before getting into overclocking and testing. The system seemed stable at 1066MHz, 5-5-5-15, and 1.8V, but not always for long. The first thing I decided to try wound up being the solution, and all it needed was a little more voltage. While 1.9V solved the problem 100%, I proceeded with testing at 2.0V, just to make sure.
With that one extra tweak out of the way, it was time to see how high this kit could overclock. The ASUS P5KC isn't regarded as the best overclocking motherboard out there, but it has helped me get to over 1600MHz on DDR3, and to 1200MHz on DDR2. It just requires a bit of patience, and some research. Like knowing not to use anything beyond the 0903 BIOS, and knowing to use the black DDR2 slots instead of the yellow slots for memory rated at 1066MHz and beyond.
As I began to increase the memory frequency I met with instability at around 1100MHz. This speed was achieved by increasing the FSB to 344MHz with the FSB:DRAM ratio set to 5:8 (the default 333MHz FSB results in a DRAM frequency of 533Mhz or 1066MHZ DDR2). The solution to the stability problem was to add even more voltage and to relax the timings a bit! Timings of 5-6-6-18 and 2.2V solved the problem, but I was now getting into voltages high enough to make me a little uncomfortable. At 2.2V I was able to push it up to about 1111MHz (347MHz FSB) before things started getting sketchy. I then went back into the BIOS and switched to a FSB:DRAM ratio of 3:5 which allows the 1111MHz DDR2 memory frequency to be maintained at the default FSB of 333MHz. The system was stable in this configuration, but didn't seem capable of much more. Even going to 2.3V and looser timings didn't help, and I was satisfied to call 1111MHz the maximum overclock for this kit.
Testing:
The test system listed in the "Configuration" section was used for the execution of all benchmarks. I wasn't sure what memory should be pitted against this 1GB kit, since every kit I have on hand is 2GB in capacity, and most of it is now of the DDR3 variety. Way back in my closet I found an old Kingston HyperX kit which seemed like a fair point of reference. This 2x 512MB kit is rated at 800MHz at timings of 5-5-5-15. What we will be able to see from this match up is what benefits can be reaped by upgrading a stock DDR2 800MHz system to include a kit of this Aeneon DDR2 1066MHz memory.
Testing will consist of four benchmarks used to gauge the performance of four different memory configurations. The benchmarks executed include tests from these three packages:
» Lavalys Everest Ultimate Edition 4.20
» PassMark Performance Test v6.1 Memory
» FutureMark PCMark05 Memory
» FutureMark 3DMark06
The four memory configurations include the following:
» Aeneon PC2-8500 DDR2 at 800MHz, 4-5-5-18, 1.8V
» Aeneon PC2-8500 DDR2 at 1066MHz, 5-5-5-15, 2.0V
» Aeneon PC2-8500 DDR2 at 1111MHz, 5-6-6-18, 2.2V
» Kingston PC2-6400 DDR2 at 800MHz, 5-5-5-15, 1.8V
These configurations represent the Aeneon memory running in its "JEDEC #1" profile at 800MHz, at the default speed and timings at 1066MHz, and at the maximum overclock of 1111MHz. The Kingston kit was run in its default configuration at 800MHz to provide a baseline for comparison to the Aeneon kit at its various speeds.
|
|
