System Benchmarks
LAME MP3 encoding will kick off our CPU-level analysis.
Within a margin of error of each other, the Time ends up a second quicker on the graphs. So far, so good. Quake3 next. While we shouldn't be limited by the graphics card, in the case of the MR9600, at our 1024x768 HQ test setting, it's going to eat into performance a little.
The analysis that 9800XT has twice the rendering horsepower as the MR9600 in the Time rings true, we'll see more of that on upcoming pages. Kribibench now.
DDR333 and higher latency memory means that the Time falls a little behind the K8V base result, to the tune of nearly 18%. With a near 18% difference in available memory bandwidth, it's clear what performance metric we're scaling with.
We're less memory limited with Realstorm, but again our slight performance deficit rears its unpretty head. The real test is in the performance when combined with MR9600.
