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.

[JoschaWondraschek]

Hi @Anonymous ,

I'm currently in the same situation of configuring a workstation.

Inventor is mainly powered by CPU speed, so I would recommend you to go with the i9-10900. ("K"-Version if you are thinking of overclocking).

As I asked in a previous post in this thread, I did some testing and comparison of different RAM clock speeds and latency times.

(I currently work on a custom Workstation with i9-9900K and have a basic RAM with 2666MHz clock speed, Latency of CL/CAL=19 and 32GB, GPU: Nvidia P1000 4GB)

I tested two different RAMs with 3200MHz (CL/CAL=14) and 3600MHz (CL/CAS=16) compared to the basic RAM and experienced pretty good performance improvements.

So in your case I would recommend you to go with the 3200MHz RAM with lower latency.

The GPU is not very important for Inventor. I sometimes work with assemblies up to 100.000 parts and my P1000 with 4GB is only about 75% full; and RAM only at about 75%. (but if you use a lot of textures in your appearance assets then I would go with a little more GPU memory)

IMO you don't need an RTX-whatever, better use the saved money and invest in some good CPU cooling and do some overclocking as you can squeeze every last drip of performance out of your CPU 😉

If i find some time I will evaluate the excel sheet i got out of the testing and will post my results if anyone is interested...

I hope I could help you with your decision making.

Let us know about your configuration and the InventorBench results 😁

[JoschaWondraschek]

Hi @Raider_71 ,

does the new InventorBench also include CAD-export (i.e. to STEP) and rendering with ray-tracing on?

[Timothy.Ward2]

I wonder if anyone has managed to get to try any of the new 5900X AMD CPU's into an inventor PC. I have half of the bits for my new workstation/gaming PC. I was planning on using this CPU in my new computer but as I was 870 odd in the queue from SCAN.COM I cancelled my preorder when I saw they were getting 24 a week. I will wait till availability is better to reorder.

[JoschaWondraschek]

@Timothy.Ward2  I just wanted to ask exact the same question 😅

I would be so interested in the performance of the new AMD processors with the faster Zen3 structure and how it affects Inventor. Especially the AMD Ryzen9 5950X.

Please let there be someone who got one before all the gamers 😂

[Neil_Cross]

Hi @JoschaWondraschek not sure about the new InventorBench as there's not much detail yet, but I can confirm that my new InvMark test does have both an export to STP and a ray tracing test.

It's not released to the public yet, we're still working on the back end website & leaderboard, optimising the scoring algorithm and finishing up user interface presentation.

I'm having a similar experience with AMD CPUs, I ordered a 5600X on launch day and I'm 370+ in the queue, I was 480+ in the queue nearly a month ago.  It's an absolute farce and now you've reminded me I'll be cancelling my order.

[JoschaWondraschek]

@Neil_Cross

Just for my understanding: You've developed a Inventor benchmarking program yourself besides InventorBench?

If it's like that, will there be some kind of "score board" where everybody can submit his results and configuration? As it would be pretty helpful for easy comparison.

[Neil_Cross]

Yes there will be an online leaderboard for comparisons which has filters and stats for comparing against other submissions, I haven't released details of that yet but here's a sneak:

In the next few days I'm hoping to release a demo of the actual leaderboard and how I've designed that to work.

[JoschaWondraschek]

@Neil_Cross

looking forward to that. Definitely giving you feedback for development 😁

[Raider_71]

Hi @JoschaWondraschek 

Yes, there will be a STEP export and Ray trace test.

I think Neil is a bit further ahead in development but we are getting some really cool early results from the new InventorBench so I am excited to get it out ASAP. We will begin beta testing next week and we have a very prominent CAD workstation supplier that will help with some early testing also. We still need to get our online leaderboard going which should be done in Jan 2021.

[Anonymous]

Computer specifications are not my
thing; maybe this is a stupid question.
Is processor clock speed divided over
the number of cores, or does each core
process at the clock speed. In other
words, a Xeon processor with a base
speed of 4.1Ghz divided by six cores
would be .68Ghz per core; or does each
core run at 4.1Ghz?

##- Please type your reply above this
line -##

[tgregory3G7FA]

It's per core.  Each core would run at  4.1Ghz (more or less).  There is variance between each core. 

You can use a program like HWMonitor to view each core.

This is my Ryzen 5 3600. (Base Clock 3.6GHz  Max Boost Up to 4.2GHz)

[Anonymous]

There is also potential clock versus net effective. Boost clock is not guaranteed. Both AMD and Intel have caveats that boost clock depends on power and cooling. They have no idea what the conditions of the system the processors will be in. Without adequate cooling, the processors will not reach boost and or maintain for any time worth talking about. Additionally with Intel, there is turbo 3.0 boost speed and turbo boost speed. One is the speed that a single core can boost to and the other is the speed several cores can boost too. There are other boosts and factors, but clock speeds are more complicated than simply two numbers.  . Also, motherboards come with default settings to allow for “c-state to be on”. Unless you turn this off, even with an overclock, your processor will throttle down and up. If you turn off c-state, you still have to go into windows and modify windows power settings, as windows provides an additional level of processor power control. Unless you have an all core clock speed, with c-state turned off, and the windows power settings turned all the way up, do you actually know exactly what your clock speed is. I know this is a lot, but just wanted to share how complicated clock speed can be. Clocks are really important and especially for inventor. Most manufacturers don’t want you to know these things which is why Neil’s Benchmark is so important. It allows us to find out what really drives the performance. . 

[bwatson1967]

@Neil_Cross Any reason why inventor bench will register a consistent 17.5-18  until I install the latest update then drops 2 points. Seems like we were great this version in performance but the update takes some of that back away.

[Neil_Cross]

---

@bwatson1967 wrote:

@Neil_Cross Any reason why inventor bench will register a consistent 17.5-18  until I install the latest update then drops 2 points. Seems like we were great this version in performance but the update takes some of that back away.

---

Which version to which version? 2020 to 2021?

I didn't notice anything unusual, but tbh I've not used InventorBench much at all over the course of this year as I've been developing my test so I wouldn't have spotted this if it did happen.

My initial thoughts are that the majority of us "cheated" InventorBench pre-2021 by hacking that registry key to increase the graphics frame rate, I know they're constantly making adjustments to the graphics engine so maybe that registry key doesn't quite have the effect it used to have? Not sure.

[bwatson1967]

@Neil_Cross I'm using 2021 and if I use the default install with no updates its 17.5-18 if I install the updates it drops 2 points. If the graphic hack you are talking about is the multithreading reg change then yes I did do that I'm not aware of any other reg change.

[JoschaWondraschek]

@bwatson1967quick question. What graphic hack are you talking about? Being curious, haven't heard of that yet.

[Raider_71]

Hi we were just toying with the idea of creating a CPU and GPU burn-in test routine. Basically, you would set the number of minutes to run the CPU (using Inventor Ray Tracing) or the GPU test to either check on overheating issues with the current cooling system in use. 

We have seen some crazy CPU throttling happenning and a routine like this could help with profiling different builds if you can visualise MHz and Temp over time on a graph perhaps.

Would this be a very useful thing for many?

[Timothy.Ward2]

It could be a handy tool. Saves downloading 3dMark/FurMark/Superposition for GPU and Aida64/Prime95 for CPU

[Mario-Villada]

Yep, that would be very useful.

perhaps consider something easy to understand for users that are not PC builders, I am thinking showing an upper limit and clearly indicate when your CPU is throttling.

Thank you @Raider_71  for such valuable contribution.

[Patrick_Gamber]

For a point of reference, updated IPI with a 3070.

Consistently scored 13.9 ish with an RX590, 3700x, 1TB Gen4 NVME.

upgraded to RTX 3070, no other changes. 10+ runs and average run was 14.25. so there is a benefit but as discussed previously its not at the same scale as what the actual card upgrade is. 

I also allow the PC to fold in the off time, RX590 would generate about 340,000 folding points per day, the 3070 is getting 3.2 Million.

ETA, this is a gaming PC that i also use as my main Inventor machine now that we are all at home. 

Our work supplied Dell SFF Core I7-6700 machines struggle to push 7.00 (bad thermals galor)