logo
NOTICE:  This is the new PunchCAD forum. You should have received an email with your new password around August 27, 2014. If you did not, or would like it reset, simply use the Lost Password feature, and enter Answer as the security answer.
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Options
Go to last post Go to first unread
jtfx  
#1 Posted : Monday, March 30, 2009 11:08:45 PM(UTC)
jtfx

Rank: Member

Joined: 3/16/2009(UTC)
Posts: 34

Model to sheet seems to be taking a llllooooonnngggg time.

I opened up activity monitor (this is on a new "nethalum" 8 core mac pro) and what I saw, I was not expecting.

The list of my processes showed viacad an 100.00% CPU, which is what I was expecting, but the little cpu usage graph showed hardly any usage (what I wasn't expecting).

So I did some research and found that the little cpu usage graphic can be expanded by double clicking on it.

Voila, I think, now I have a chart of CPU usage by cores, but to my surprise there are 16 shown, I was expecting 8, and only 1 (what I would call core 2 based on the graph)is doing anything?

Is there a preference I am missing that could be changed to take advantage of more cores?

Right now I have the famous spinning beachball while generating an aux. view in sheet mode so I can't check viacads preferences.

The system preferences doesn't seem to have any related to processors.

Any thoughts?
nick  
#2 Posted : Tuesday, March 31, 2009 2:51:06 AM(UTC)
nick

Rank: Senior Member

Joined: 2/2/2007(UTC)
Posts: 245

Hello JTFX

As we use ACIS as our modelling kernel and ACIS is not currently multithreaded we are unable to use more than a single core for any modelling function. If you had Shark FX you would see when you rendered a file that LightWorks is multithreaded and we take advantage of all cores when rendering.

Hope this answers your query.

All the best
Nick
jtfx  
#3 Posted : Tuesday, March 31, 2009 6:14:03 AM(UTC)
jtfx

Rank: Member

Joined: 3/16/2009(UTC)
Posts: 34

Nick,

Thanks for the quick response.

I am new at this and aren't sure when a file "renders" vs other machinations.

Would model to sheet be faster in Shark FX, or only the fancy shaded views would be faster?
posh.de  
#4 Posted : Tuesday, March 31, 2009 6:36:29 AM(UTC)
posh.de

Rank: Senior Member

Joined: 2/23/2007(UTC)
Posts: 484
Germany

Thanks: 28 times
Was thanked: 50 time(s) in 36 post(s)
Originally Posted by: jtfx Go to Quoted Post
...but to my surprise there are 16 shown, I was expecting 8, ...


each of the physical processors are divided in two logical processors by the [URL="http://en.wikipedia.org/wiki/HyperThreading"]intel Hyperthreading technology[/URL].


Originally Posted by: jtfx Go to Quoted Post
and only 1 (what I would call core 2 based on the graph)is doing anything?


currently all modeling kernels are using one (1) kernel only because slicing of geometric calculations is much more difficult than the rendering of 'dumb' raster data in multiple slices as done by multi-threaded render & raytracing apps.

therefore the fastest system fr 3D CAx apps you can currently get does use a dual-core CPU as high clocked as possible (as e.g. an intel Core 2 Duo Extreme) instead of multiple lower clocked quad-core CPUs twiddling thumbs most of the time.

this may/will change for the future of course.

hth,
Norbert
nick  
#5 Posted : Tuesday, March 31, 2009 6:49:19 AM(UTC)
nick

Rank: Senior Member

Joined: 2/2/2007(UTC)
Posts: 245

Hello JTFX

No model to sheet speed would be identical in Shark FX. As per previous post only task accelerated by multiple cores is photorealistic rendering - http://en.wikipedia.org/wiki/Re...ring_(computer_graphics)

All the best
Nick
jtfx  
#6 Posted : Tuesday, March 31, 2009 8:45:14 PM(UTC)
jtfx

Rank: Member

Joined: 3/16/2009(UTC)
Posts: 34

Originally Posted by: posh.de Go to Quoted Post
each of the physical processors are divided in two logical processors by the [URL="http://en.wikipedia.org/wiki/HyperThreading"]intel Hyperthreading technology[/URL].




currently all modeling kernels are using one (1) kernel only because slicing of geometric calculations is much more difficult than the rendering of 'dumb' raster data in multiple slices as done by multi-threaded render & raytracing apps.

therefore the fastest system fr 3D CAx apps you can currently get does use a dual-core CPU as high clocked as possible (as e.g. an intel Core 2 Duo Extreme) instead of multiple lower clocked quad-core CPUs twiddling thumbs most of the time.

this may/will change for the future of course.

hth,
Norbert


Thanks for the information, maybe my duo core laptop will be faster then my desktop, lol? Kind of sad actually...

My old dual 1 GHZ G4 processor "Quicksilver" machine seems to use both processors, well at least the little green bars appear to be going up and down approximately together,
Users browsing this topic
Guest
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.