|10-15-04, 10:02 AM||#1|
Join Date: Oct 2004
Location: Milan, Italy
Gaphs: DST vs no-DST & SM 3.0 vs SM 2.0 & full precision vs partial
Graphs that show difference between:
1) DST vs no-DST http://www.hwupgrade.it/articoli/1094/17.html
2) SM 3.0 vs SM 2.0 http://www.hwupgrade.it/articoli/1094/18.html
3) Full precision (32 bit) vs partial precison (32&16 bit) http://www.hwupgrade.it/articoli/1094/19.html
-The conclusions are:
1) There is an huge difference between scores with DST active where NV4x gpu gains around 10% points over a no-DST score.
2) SM 3.0 seems much less important because also if in the graphs the diffenrece between 3.0 and 2.0 seems huge we have to keep in mind that in 2.0 3dmark05 deactive the DST, so 90% of the loosed points are because of DST not active.
3) Full precision seems going from just slower to much slower depending on the situation; while nv30 suffer like hell using Full precision, nv4x shows to be much stronger and suffer much just in the pixel shader test.
Support GeForce FX owners againts the unfair business of ATI & Valve vote for this petition!!!