DDR2 results in faster sound benchmark compared to DDR1
Can someone explain why I get better results in 3D Mark 03's sound
benchmark when I use DDR2 RAM instead of DDR1?
The mainboard can take both DDR1 and DDR2 (not at the same time though).
I've run the test with DDR1 (@200 MHz, timings 3-3-3-7 2T) a few
times. Then with DDR2 (@266 MHz, timings 4-4-4-12 2T). Every single
time the sound test was much faster with DDR2.
FSB was the same each time (271MHz). The FSB/DDR ratio is 1:1 in the
case of DDR2, and 4:3 for DDR1. RAM timings weren't "pushed"; the
modules are rated for those timings.
Re: DDR2 results in faster sound benchmark compared to DDR1
On Sat, 29 Mar 2008 11:40:46 +0200, Nikos Chantziaras
<realnc@arcor.de> wrote:
>Can someone explain why I get better results in 3D Mark 03's sound
>benchmark when I use DDR2 RAM instead of DDR1?
>
<SNIP>
>I'm scratching my head why I get better results in a test that has
>nothing to do with RAM whatsoever ?:/
It does have something to do with ram. All parts of the
system work together, but where does the CPU get the data it
processes? What happens if it can't get the data fast
enough? Memory is always a "potential" bottleneck but often
fast enough it doesn't reveal itself as one before some
other part is a more significant bottleneck.
Sound benchmarking is about processing sound, what your
system has to do before sound data gets to the sound card
output, not about what the sound card outputs to the
speakers. Memory is bottlenecking the CPU's performance and
you can also see the result in the CPU test on your linked
pics.
DDR2 results in faster sound benchmark compared to DDR1 
Linear Mode
