Synthetic Benchmarks:

Our objective here is not to break the sound barrier or hit the speed of light, but we are going to demonstrate the capability of our test configuration according to real world testing and using synthetic API modules to test performance characteristics. Let's have a look.

FSB x (17.0)	VPU - RAM Clock	Aquamark 3 Score
200 MHz	400MHz/500MHz	61,451
220 MHz	400MHz/500MHz	64,701
233 MHz	425MHz/525MHz	69,192

Our system's CPU was juiced up from 1.325 ~ 1.400 volts, just for reasons of stability, not that we encountered any, however I wanted to be sure this configuration will be stable under gaming loads and benchmarks. On the flip side of things we can clearly we can see that the PowerColor X800XL puts out some nice scores and can definitely hold its own. This card is looking real good!

FSB x (17.0)	VPU - RAM Clock	3DMark05 Score
230 MHz	428MHz/533MHz	4889

For 3DMark05 I bumped up the VPU/Memory core to 428MHz/533MHz and put the 17.0 CPU Multiplier to a 230 MHz FSB and pulled off this score with about 30 processes running in the background. This allows us to get the real world results we are looking for. We probably could of ripped off a better score with some extra tweaking but this is all we need to see as a base for testing. You can expect similar scores with a similar system. Let's see what Sandra 2005 has to say about the Radeon.

Sandra 2005 Professional

The floating point and integer benches pretty much test how much binary can be processed at the VPU/CPU exchange level. What this means is we are seeing how well the mother-board CPU and VPU of the X800XL is communicating by processing Direct 3D in a synthetic multimedia scenario.