Hey, I was wondering what sort of performance people get out of their graphics card in blender before you notice slowdown. I don’t know the best way of figuring that out - maybe adding the default cube and continuing to press subdivide till you notice jumps when you rotate? Or maybe a test in sculpt mode?
I’m still not so clear as to just how much difference there will be between say a 7600gt and a 7950gt when it comes to workflow, as in amount of polys displayed before it becomes hard to move around. I would like to figure this out before I make a final decision on the card to get. I may trade my card specs down and instead go for an extra 2 gigs of ram. I’m still on a laptop with integrated graphics, and I get 25000 verts and it becomes choppy.
Maybe we can post our results like this:
Processor: 1.6ghz centrino
Ram: 512mb
Gfx: intel gma crappy thing
Workable vert limit: 25000
1,5 years old sony vaio laptop, win xp.
cpu: intel pentium m74O, 1,7ghz.
memory: 512mb (2mb L2 cache)
graphics: 128mb nvidia geforce go 64OO
workable vert count: 9O.OOO //aint enough 4 nothing, i should have 1,5 million to actualy enjoy the real benefits of having the sculpt mode, so for this i should have at least 4 gig of ram //
add up one more multires level:34O.OOO -still manageable although very sluggish, but within a few minutes and after a few operations i get a crash.
It starts to slow down at a million polygons but a lot of the systems mentioned above should be able to push that many polygons in an optimized OpenGL renderer pipeline.
System:
Ram: 4GIG (3.25GB visible) DDR2 800MHz, 4 1GIG modules paired up - same make and model
Motherboard: Intel 965
Graphics: GeForce 8800GTX
CPU: Intel Q6600 Core 2 Quad @ 2.4ghz
