ZBrushCentral

E6850 Dual Core vs Q6600 Quad Core

Hi. I’m building a new home computer.

The main question that remains unanswered is which CPU to choose, and, with games currently GPU limited, it will be ZBrush that becomes the decisive factor.
I’ve already settled on an Intel (my first in many years). After the recent price cuts, I’m doubting between two of them: E6850 and Q6600, which cost the same.
While rendering will surely be faster in the latter, I was wondering which one would work best in day to day work with Z3: higher clock speed (2.93 vs 2.33 GHz), or more cores?

I think that the additional cores will provide a better improvement in performance than faster clock times will. Preferences>Performance>MultiDraw applies specifically to editing your mesh. So ZBrush will take advantage of your extra processors to give you better performance when sculpting.

aurick, this is great news indeed! because, previously i’d give an answer almost instantly that higher clock speed will give you a better boost with zbrush… cool! especially because i plan a 8core purchase in near future :slight_smile:

Nekh, additional good news is that Q6600 is very easily overclocked… without any fuss it can go to 2.9ghz!

That settles it then. Thanks for the info. :slight_smile:

I just upgraded from a dual to a quad a few days ago and its definitely faster. Preview shadows render much faster now too.

can you tell us your system specs and with how many polygons you are able to comfortably work?

Lets see…

Vista Home Premium 64bit
EVGA 680i Motherboard
4gb RAM
qx6600 CPU
Nvidia 8800 GTS

Honestly I havent really run into an upper poly limit yet. The figure I’m working on now is up to about 7.5 million and still runs smoothly. All I can say is that I turned off preview shadows with the dual core and I leave then on now.

hm… i’m working on a clocked E6600 running at 2.9ghz and 2gig RAM at 800mhz and xp pro and i can work with 8.5mil polygons with preview shadows on smoothly (occasional mem compacting… but that’s not a cpu issue)… your machine has to be able to chew more than that smoothly…

it would be really useful thing to know, so please let us know if you get a chance to run more thorough tests

hmm

i am workin on a e6600 4 gig 2x 8800GTS SLI winxp 2xNEC 21" LCD
(sux) (even with sli and one LCD)

home i own a qx6700, 4gig, GTX8800 786mb, vista ulti 32, ageia physx, 47" Panasonic Plasma
(rox)

well the 4 cores are in my experience much more better to have IF the application OR the game support THAT multicore, if not… waste of power.

game eg: lost planet, where you can decide how much procs are used to calculate… if i decide to use 2, game get 30fps less…

rotatin an object with 40somethin mio polys with XP stutters, with my vista box aint. (i dont think the ageia is doin somethin with zbrush :wink: )

mathematically we dont need to talk about, because everything that counts is your workflow and how YOU feel nice.

exept a well known TDR-bug in vista i got more fun workin a home :wink:

ps on both machines:
Compact mem 1400
both undos to 4
MaxPolyMesh is set to 50

I’m thinking that I might add xp to my new system (as a dual boot). I’m using vista now, and I’m having a lot of little glitches and a few big ones that I can’t really live with while I wait for service packs and better video and wacom drivers etc… Does anyone here know if xp utilizes the quad core chips very well - esp. for use with zbrush?

hi,

on the upper floors we are using the quad core machines for post and videoproduction… theese are running with winXP and winXP64 (because of to take advantage of the 3GB+ ram)

they run smooth and easy… i dunno how it is with zb, but might be nice.i can ask one of the users to let me install my versions of zb.

what kinds of difficults you are running trough with vista? maybe i can help you.

me got an intuos3 and teh cintiq 21UX with vista… no problems at all :\

regARTs
k0s

edit:

as aurick mentioned in a thread before, it makes sense to place the Multithreading buttons on the UI or use the script, where zb initializes with multithreading feature.
personally i am using that script and placed the high-prio button on my shelf. as i said works well… exept the wk bugs :wink:

I have the following:

EVGA 680i mainboard
4GB Corsair dominator ram
E6300 cpu overclocked to 3.1ghz
Windows Vista Ultimate 64bit
8800 GTX

I can currently handle 20 million polys without any major problems.

For 3D Graphics memory is crucial as well as a decent GFX card. My e6300 is a nice CPU although I will soon be upgrading to a quad core because they are the future and software will soon make then a must.

as i understand, zbrush currently does not use 3d acceleration…

but wow! 20mil… nice! :slight_smile: i should really upgrade to 64bit sys

rotatin an object with 40somethin mio polys with XP stutters, with my vista box aint.
^^ :wink:

you should :smiley:

atm i have experienced that zbrush got his end with 50mio polys… when i am with 41mio on the road i cant subdivide anymore, because it says: check mem:pref settings and set teh maximum allowed polys higher… well 50 is the end of the range :smiley:

just to clear out things, i am not talking about the HD divisions :wink:

you guys are talking some really nice numbers… i most definitely should! :slight_smile: and when i remember that i thought 8.5 was cool hehe

if i wasn’t in the middle of the project i would probably start installing right away :)…

Yep it’s nice to have access to that 4 Gig. When I say 20 mil that is the furthest I have taken it so far and I haven’t even tried to get past that so I’m sure it can go further.

In fact I will try now to see how far I can get it… 61.5 million polygons using the demo head and adding multiple zshperes to it and sub dividing each to its max. At the 61.5 mil mark my PC started to slow down and Zbrush use became impractical. At the 50 mil mark though, it was still usable.