How fast is your Inventor PC really?

[Raider_71]

Hi guys,

We have had to do some testing on a bunch of Inventor PC's recently to determine which of the PC's needs to be replaced. Obviously we needed to find out which of the PC's are the worst eprformers as there was only budget to replace 50% of the design PC's. So we thought the Darwin theory will come in handy right... 🙂

Anyways I started searching on the net for toppics on how to benchmark an Inventor PC. Then I thought whats the point of using gaming benchmark tools because Inventor is not a game and there are more aspects than just graphics performance when it comes to percieved performance on an Inventor PC right.

So we decided to create our own Inventor benchmark tool which tests various aspects of an Inventor PC to give us an overview of our PC's performance. This then helped us make a decision as to whcih pc's to replace.

We have made the tool available free of charge to anyone interested in checking how their PC stacks up to their peers or friends. 🙂

Please download it here and post your results here as well if you want. Would be interesting to see what beast workstations are out there.

I would like to say thanks to Kirk #karthur1, for helping in testing the app.

Please feel free to send any suggestions our way. There is an email link in the app.

Download and Install

The application will work with Inventor 2014 to 2016 only.

IMPORTANT: After installation there will be an Inventor Bench icon on your desktop that looks like this: 

My resluts:

HP Elitebook 8560w with an SSD upgrade.

[Neil_Cross]

There's much much bigger assemblies out there bigger than a 2267 unique occurrence count.  It subjectively falls under the definition of large assembly, but it's by no means massive.  Inventor is very capable of handling assemblies of that size, I wouldn't agree with saying Inventor isn't designed to be pushed that hard, I see assemblies every day bigger than that with no issues.

The way I explain it to people is that when Inventor is working with thousands of parts, it's handling thousands of physical files, therefore imagine if you asked MS Word to open thousands of busy documents into one session... it's not necessarily technically the same but it's the same in principal.  Drawings can be a little bit different, if you're displaying 110,000 components on screen and asking the machine to calculate line details with depth calculations for hidden geometry, it's a ridiculously big task.  I think Inventor could still handle these situations better but either way at ~2200 active documents it isn't pushing the limits of the software.  What problems are you having? (might be better starting a different forum post for that)

[smokes2998]

In general no matter what the hardware inventor runs on, It is a lot slower than SE and NX at handling large assemblies and it takes a long time to load  and update.

This is due to SE and NX having better methodology of what to load in main assembly.

with drawings,

I have had days when it took 8hr to update a drawing.

It more an issue with the fact that the inventor loads the 3d model into the drawing rather than creating a vector image like SE.

[mpatchus]

---

@smokes2998 wrote:

I would really like to see how user rate the performance of the software when handing assemblies this big

and drawings  this big this a 22 sheet a2  drawing of the assembly

when you do everything right in terms modeling and assembling you will find the software is the bottleneck as it isn't design to be pushed this hard

---

A typical average size tie station that we regularly design.

Everything included right down to the nuts & bolts, and of course all of our cable runs done in routed systems.

I have not encountered the bottleneck you describe, even on our archaic Lenovo P300 machines.

[machiel.veldkamp]

@Anonymous Very interesting indeed.

One might say to use substitutes in your assembly to ease the workload.

As a counter to that argument: Inventors Substitue function does not work as intended with large assemblies.

Yes in theory it sounds great and when people show it off it seems that it works fine.
HOWEVER> I find that in a working model (i.e. a model that is activly being worked on by multiple engineers and is thus changed often) this function breaks as soon as a bottom assembly is changed. The substituted model WILL not update.

So. Just wanted to let that small rant out.

Any thoughts, mpatchus?

[mpatchus]

We have made use of substitute LOD parts for much of our structure assemblies, which rarely change, and are stored in our library for designers to insert as they require.

Once we got them to remember to do a "Link Levels of Detail" within their station models, 99% of the "problems" they were seeing went away.

You can realize significant gains by using substitute LODs.

Good example is one of our standard lattice columns.

Normal column assembly... 992/44 instances

Substitute assembly ... 281/15 instances

Since we use upwards of 30 columns in a station, you can begin to see the implications.

We suppress all of the hardware, but it isn't just the number of instances that is impacted, we also eliminate 95% of the holes in the parts.

Thus less surfaces to be calculated come drawing time.

Face it, on an 80' tall column on a "D" size sheet of paper, are you going to see more than a speck for a 1/2" bolt head?

[machiel.veldkamp]

"which rarely change"

That's the thing isn't it?

You know. You just gave me an idea. We have a bunch of assemblies in our Vault. I can make those in Subs! I guess every part counts.

[Anonymous]

the test run with Inventor 2017 testversion, after full installation and running on default settings.

[smokes2998]

Ever done and multi-sheet assembly drawing with exploded views of each assembly step? 

This where Inventor get really slow especially on the updates.

[proj964]

Not to put too fine a point on it, but isn't this thread to work on what hardware provides the best Inventor performance given iinventor is what it is?

[machiel.veldkamp]

Yes. But understanding what makes some assemblies slow is part of choosing hardware I think.

[blair]

Using Sub Assemblies helps with performance as well. 2905 / 319 instances:

[Anonymous]

http://zupimages.net/viewer.php?id=16/29/c09m.png

Nvidia Maximus with Quadro

& Tesla no difference.

[Neil_Cross]

I must admit I've zero experience with Teslas, never even looked them up online until now so I've no idea what I'm talking about when it comes to them.

But is that a very old Tesla? The specs are average by todays gaming cards standards?

[smokes2998]

If you use CPUID you will see where the bottle necks are when loading a large assembly.

 I would like to see how much faster a gaming card will make as I run a quadro k4000. with an SSD and 32gb ram 3.5ghz i7. I think the performance is slow compared to competitors software as they can lightweight a large assembly much much better than Inventor.

[mdavis22569]

Old test

@mdavis22569 wrote:

which is odd .... with a 6700 

Thinking it's my old video card 

new test 

@Neil_Cross

[machiel.veldkamp]

That is impressive. DO you have that CPU running at factory settings or have you overclocked it?

What kind of cooling do you run?

[mdavis22569]

Super Cooler and nope' I haven't Overclocked it yet .. 

That's all out of the box settings

[Neil_Cross]

I don't think there's many surprises left to be had here and anybody who's followed this from day 1 will probably now know what to expect, but with that being said, I've just done a very interesting test which doesn't throw up any surprises but is still interesting to see.

Early test with a 4.8GHz CPU (before the numerical rating was introduced) the CPU was overclocked and you can't see that in the info

Test done today with a CPU at 2.4GHz (it should boost to 2.6GHz but HWMonitor reported all cores running at 2.4)

Exactly the same graphics card in both tests, dramatically different visual performance results.

More interesting though is the totals.

Model save time = 5.41s VS 10.06s

Modeling test = 10.28s VS 21.98s

Drawing test = 22.3s VS 45s

CPU = 4.8GHz VS 2.4GHz

Exactly double the GHz almost exactly double the performance, almost to the second.

Co-incidence? Possibly.  But you can't take a CPU out of one system and just slot it into another.  It's definitely not an apples for apples test but I found this to be a very interesting set of results.

[Anonymous]

But maybe anyone has the chance to underclock a cpu and deactivate the turbo modes of a system. Then we could see the developement of the performance at 1.5 / 2.0 / 2.5 / 3.0 / 3.5 / 4.0 GHz or any other variation of clock speed. All other hardware and software specs would be exact the same.

[mdavis22569]

I did turn off one more program and got to 11.39.   I'm tempted to overclock both the cpu and video card .... I could get about 4.8 on the chip .. and put the card into O/C or gaming mode. 

I also picked up a LG 29" Ultra wide monitor yesterday..