Concerning the 3DMark score. I have discovered something that really bugs me about this number.
It is calculated by simply adding the 6 game simulations together and multiplying by 12.
Sounds allright up front, but here is a scenario that would make the numbers bogus.
I will show 2 scores side-by-side.
Low Detail Helicopter: 20 15
Medium Detail Helicopter: 15 15
High Detail Helicopter: 10 15
Low Detail Adventure: 20 15
Medium Detail Adventure: 15 15
High Detail Adventure: 10 15
The numbers are bogus and just for illustrative purposes. The two systems would have the exact same 3DMark score. Is it possible? Probably not to the extent I show, but it is possible.
I actually ran into a scenario during all my benchmarks where 2 very different video cards ended up with a very close 3DMark score. Its something similar to what I show above. One card would just smoke at low detail settings, but then barf at the high detail settings, while another ran so-so on low detail, but really did not slow down much at higher detail settings.
I got to thinking about it, and setup my own BDPoint system which gives weighting to the higher detail tests. After all, if a card can smoke through those, then it should get more points.
So, I am re-working the WEB site, again
and adding my point system along with the 3DMarks.
I hope to have the first series done for review and comments in a couple of days. Then I will put up everything after that.
------------------
Roy "Skuzzy" Neese
President, AppLink Corp.
http://www.applink.netskuzzy@applink.net
