3DMark 06 Fill Rate:
Excellent scaling for CrossFire here! The dual X1900 XT was actually a hair more than twice as fast as a single X1900 XT.
3DMark 06 Perlin Noise (SM3):
Again, we have just a hair under 2x the speed of a single X1900 XT when running a pair of them with Crossfire.
3DMark 06 Pixel Shader Performance:
Ok, we begin to see a clear pattern here... Crossfire with two X1900 XT's is twice as fast once again.
3DMark 06 Vertex Shader:
The pattern is somewhat broken. For complex vertex shaders, the Crossfire configuration is again twice as fast as a single X1900 XT, however very strangely for the simple vertex shader a single X1900 XT is MUCH faster than two of the cards in a Crossfire configuration! The only explanation I can think of is that the overhead of transferring the simple textures outweighs the computational benefit of Crossfire.
forums
send article 
hardware newsletter 
article comments (1)
Digg this!
del.icio.us