why dosent inventor use multi core threading processing

why dosent inventor use multi core threading processing

r_blackburn3000
Enthusiast Enthusiast
7,394 Views
10 Replies
Message 1 of 11

why dosent inventor use multi core threading processing

r_blackburn3000
Enthusiast
Enthusiast

I am running a 4 core i7 3.5 ghz with 12 gig of ram, and on a single spiral wrap in a part I have to wait

10-15 minutes to get out of sketch mode. I just dont see how more horsepower/video card would help if inventor 15

only uses 50 threads during a not responding procedure. This would not fly in industry, how does one get around it.

 

My part is a simple cable with a helical twist pitch only 8" long.

0 Likes
7,395 Views
10 Replies
Replies (10)
Message 2 of 11

mcgyvr
Consultant
Consultant

post the ipt file..

 

Some portions of Inventor are multi-core capable (FEA/Studio/IDW)

Modeling is not..

Most if not all 3d cad programs still cannot leverage multi-core for the modeling side of things..

a much smarter geek than I would know the real behind the scenes reasons.. but its something about the kernel needs to be 100% rewritten blah blah nerd blah

 

 

 

 

 



-------------------------------------------------------------------------------------------
Inventor 2023 - Dell Precision 5570

Did you find this reply helpful ? If so please use the Accept Solution button below.
Maybe buy me a beer through Venmo @mcgyvr1269
0 Likes
Message 3 of 11

dgorsman
Consultant
Consultant

Hmmm nerd senses are tingling.  Or just sleep deprived...

 

Multi-threading isn't a magic wand that can be waved at any old pumpkin to turn it into a muscle car.

 

It's pretty easy to break down.  In order for multi-threading to work on a process, that process has to be broken down into different tasks that can be worked on more or less independantly of each other.  Look at what you are doing in Inventor, break that down into gross steps, detail those out a bit further.  Which of those are dependant of each other (FEA, image processing, as noted)?  Which are not?  The former are a good candidate for multi-threading.  The latter are not as each operation must wait for another to finish before starting.

 

A more general example is operating in a very large database.  A look-up operation (match records with field content "Blah", match records with field content "More Blah") can be executed in parallel with each other since they don't affect each other.  A generic write operation cannot be strictly done in parallel, since two operations can't change the same record at the same time - one must wait.  But with some strategic organization so that the write operations must always work on unique sets of records they can be done in parallel, but it won't be the same level of improvement as the parallel read operations.

----------------------------------
If you are going to fly by the seat of your pants, expect friction burns.
"I don't know" is the beginning of knowledge, not the end.


Message 4 of 11

cbenner
Mentor
Mentor
0 Likes
Message 5 of 11

swalton
Mentor
Mentor

If you can, post your part.

 

A different modeling technique might speed things up.

 

 

Steve Walton
Did you find this post helpful? Feel free to Like this post.
Did your question get successfully answered? Then click on the ACCEPT SOLUTION button.

EESignature


Inventor 2025
Vault Professional 2025
0 Likes
Message 6 of 11

JDMather
Consultant
Consultant

r_blackburn3000 wrote: 

My part is a simple cable with a helical twist pitch only 8" long.


I already know what the problem is.  Attach your file and I will verify. (it is an error in operator logic if my guess is correct)

 

jmather_at_pct_dot_edu


-----------------------------------------------------------------------------------------
Autodesk Inventor 2019 Certified Professional
Autodesk AutoCAD 2013 Certified Professional
Certified SolidWorks Professional


0 Likes
Message 7 of 11

r_blackburn3000
Enthusiast
Enthusiast

 

 

Edited by
Discussion_Admin

0 Likes
Message 8 of 11

vex
Collaborator
Collaborator

May want to remove/edit the above information. I don't think you ment to post that.

0 Likes
Message 9 of 11

r_blackburn3000
Enthusiast
Enthusiast

Disregard the contact info, I was trying to send a file.......

0 Likes
Message 10 of 11

JDMather
Consultant
Consultant

Is it OK to post solution file here?

(almost everything in the file was done incorrectly)

or would you rather I keep this private via email contact?


-----------------------------------------------------------------------------------------
Autodesk Inventor 2019 Certified Professional
Autodesk AutoCAD 2013 Certified Professional
Certified SolidWorks Professional


0 Likes
Message 11 of 11

JDMather
Consultant
Consultant

@r_blackburn3000 wrote:

My part is a simple cable with a helical twist pitch only 8" long.


The main problems I found (and fixed)
-your conductor profile was not perpendicular to the path, this results in computationally (and incorrect) expensive geometry.

-your conductors were not separate bodies, now that computationally expensive (and incorrect) geometry just got a lot more expensive in calculating the merging of those complex eliptical, twisted faces.

 

The simplified (and more geometrically correct) geometry solves much faster.

 

Conductor.png

 

Let me know if it is OK to attach my file here as it may be instructive for others.

 

Oops: I have the twist going the wrong way.  Smiley Wink

 

 


-----------------------------------------------------------------------------------------
Autodesk Inventor 2019 Certified Professional
Autodesk AutoCAD 2013 Certified Professional
Certified SolidWorks Professional