New to Vista - Windows Experience Index weirdness

I just received my new work laptop and am in the process of setting it up. I have a quick question about this Windows Experience thing (which is new to me - my last Windows machine ran XP).

I know the quick answer is that I should probably ignore WEI entirely and get on with things, but I'm curious. Why would a "desktop graphics" score be lower than a "gaming graphics" score?

It seems weird to me because I chose this machine specifically for office duty...

WEI (before RAM and HDD upgrade)
Processor: 5.6
Memory: 5.9
Graphics: 4.0
Gaming G: 4.7
Primary HDD: 5.3

Any explanation?

 

desktop graphics (afaik) only tests raw fillrate of huge triangles (to animate and display the windows) and maybe the ram amount (to know how many windows could be saved).

game graphics measure the performance for lots of small triangles, vertex prosessing, pixel shaders, etc. and there, it looks your gpu excels, but not at the raw fillrate.

at least, this could be it.


now the posts will pop up about forgetting about WEI entirely. i don't agree with that.

 

Thanks! That would make sense. I'll recalculate WEI after I double the RAM and see if it has any effect.

As for the usefulness of WEI -- it seems like an issue that would probably be debated to death on message boards, so I'm not asking anyone to go through it again. Good to hear that some folks consider it a useful tool. I'm sure hardcore gamers prefer other benchmarks, but if WEI works, then it seems like it could be a handy built-in diagnostic...

 

IN before the "no" posts

cheers ...

 

This explains what each WEI test does to determine the score. As you can see from that article, WEI is pretty basic, and isn't a good determiner if something is really good or not. It shouldn't be discarded completely, but it's not a great measure of performance.

 

Using WEI to benchmark a system is like measuring distances with a ruler that is only divided into quarter inches (or, for the SI-inclined, a meter stick that is only divided into centimeters). You can get a reasonable approximation of the actual distance, but if the distances you are measuring are small, or if the difference you're trying to detect is smaller than the scale of the ruler, then that ruler (or meter stick) becomes useless.

So it is with WEI - it's an approximate way of determing if the gross performance criteria of one system differ substantially from those of another system, which is all most people need it for; however, it is not useful in doing accurate benchmarks because the scale it operates is too large to measure the differences that matter.