Kingston HyperX 512 DDR PC3500 - Kingston HyperX 512 DDR PC3500
(Page 3 of 4 )
3DMark 2001SE Pro
434MHz 450Mhz
With the increase from 434 (217x2) to 450 (225x2) we netted a 26 point gain, but also don't forget we went from 2170MHz down to 2138MHz on the CPU with the multiplier adjustment. (10x217 for the first score, 9.5x225 for the latter)
3DMark 2003 Pro
434MHz 450Mhz
With the jury still out on 3DMark03, we decided to throw it in to start to build a database. Contrary to the 3DMark01SE score, the increase in FSB did not help the score any. We ended up losing some points, which leads me to believe that the test is a little more CPU dependant than its predecessor. (the benchmarks were run 3 times to ensure consistency)
Pretty impressive performance running the recommended timings. Let's see if we can get it to perform with more aggressive timings. We'll start at 434 (217x2). I increased the timings to 2-2-2-6 at 2.5V. Prime95 would get errors at until the voltage was increased to 2.7V. (the cpu was prime stable @ 217x10, and the cpu was not a limiting factor.)
SiSoft Sandra Memory Benchmark
434MHz 2-4-4-8 434MHz 2-2-2-6 450MHz 2-4-4-8
With the more aggressive timings (2-2-2-6), running 434MHz, we are getting better results than recommended timings (2-4-4-8). Just to prove the point that higher FSB doesn't always mean better performance.
PCMark 2002 Pro
434MHz 2-4-4-8 434MHz 2-2-2-6 450MHz 2-4-4-8
Again, with the aggressive timings we are seeing a better performance boost than the higher FSB with the relaxed timings.
Next: Kingston HyperX 512 DDR PC3500 >>
More Memory Articles
More By Cygnus_X_1
| Recommended by Dev Hardware |
|---|
|
