@pep75 said:
I had a Nvidia 8800GTS 32OmB card and sold it to buy myself a "professional" graphic card with sufficient memory on it.
video memory is typically no issue, at least using a video card w/ 256/512mb on board (not shared w/ the OS). More memory does not improve the output speed in any way and thus is not recommendable besides other apps as e.g. raytracing programs can make use of it
@pep75 said:
The card costed 500,- euros and a 8800 gts costed 150,-. And well I can hardly see the difference between these two really. OK, the Quadro card is slighty faster, but I'm really not stunned.
the Quadro FX series provides an improved maturity and stability for the OpenGL based 3D output mainly for mid-range/high-end CAx apps and the driver optimizations available for them. The speed in connection w/ mesh-based 'standard' modelers as e.g. SU or Raytracing apps will normally not profit by using a Quadro FX, actually a GeForce in the same price range is often faster because of using a more recent GPU.
@pep75 said:
I'm working on a 64bit platform (windows 7),8gb memory, the processor is a 6600 quadcore and my hardisk is a 80gb Intel X25 ssd. Cinema4d is going like a razor but Sketchup feels like sluggish....
C4D is multi-threaded, SU currently not. Which is understandable because slicing a render task generating raster data (i.e. pixels) is much easier than distributing vector based modeling operations over several kernels.
The only way to improve the performance of SU is to use a faster CPU in the sense of speed of one (1) kernel, i.e. a faster dual core is better than a slower quad core... currently.
hth,
Norbert



