Locked at the request of the OP.

For the record, many people want to compare one score to the next as if they are on the same level.  You cannot compare DX10 to DX11 scores.  You just can't.  Let's for an instant realize one important fact - there are no comparable DX11 Nvidia cards on the market.  When the 300 series comes out, then you can compare and contrast the cards, and call one card better than the other.  However, unless someone wants to test a 5800 in Vista, in DX10, you can't compare it to a 200 series card for scores.  However, the bench from anandtech (*link*) does clearly show that it works better in DX10 than the 200 series.  In fact, quite better.  However, anandtech does attempt to trivialize the bench.


If you need a better example of what I mean, then take a look at Crysis.  Running the same settings in DX9 and DX10 mode lead to much, much different results for FPS.  To be precise, DX9 mode will have much higher, sometimes even 2-4 times higher FPS than DX10 mode.  To that extent, many of you missed luv2's point.  The reality is that until Nvidia puts out a card that can run against the 5870 in DX11 mode with this bench, comparisons between the DX10 and DX11 versions of this card are pointless, and tell you absolutely nothing between DX10 and DX11, and certainly can't tell you anything about how the 5870 will compare with the 300 series.  Once the 300 series is released, this benchmark may be come relevant again.


/thread