Community
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... ๐
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. ๐
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.
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.
I take it you're on a 32bit computer then? I have no idea how you can work with that restriction on those assembly sizes, I'd be hitting the roof to get those upgraded.
Have you investigated the 3GB switch which was used quite a lot in the old days on 32bit systems?
The switch obviously allows applications to address 3GB RAM. It's no substitute for moving into the modern world though.
We are on a 64bit system, and I have 32gb of ram installed, but the most Inventor ever uses is 2gb.
So needless to say our boxes spend a lot of time writing to the drives.
The solid state drives helped enough for modeling our substations, which are fontunately nowhere near the size of the tie stations.
I've told them no more tie stations are to be modeled until 2016 is up and running.
Ehhh what's going on there then? Inventor 2012 has a 64bit version so it should be able to use all of that RAM, there are no default limitations stopping that from happening, has someone within your company intentionally prevented Inventor from accessing more than 2GB RAM?
Anyone run or know anything about "drive accelerators" like the Fusion iodrive ii
We just picked up 5 at an auction and IT only intends to use 3 in our servers so IT has mentioned putting one in my computer instead of a regular SSD..
IF they do I will benchmark but just wondering if anyone uses them now..
These are the ones we got..
https://www.corporatearmor.com/product_info.php?products_id=11402&gclid=CNKOhKGi18sCFQIfhgodKdwFEw
I think they said they got them for $500 each ($13,999 MSRP 
)...
That thing looks lethal, I'd never heard of them however after reading up on them, it looks like they're only really going to excel in server platforms... by the looks of it you can get similar read and write speeds through existing much cheaper high end consumer SSD's but these things have ridiculous off the charts IOPS capabilities. Inventor or even the desktop in general just doesn't write to the disk frequently enough to make use of something like that but it'll be fun trying!
Can't say that I have but my MacPro uses SSD but through the Pcie interface and I'm just putting together a new system for work next month that uses simular with the SSD connected through the PCie interface.
I'll bench it next month when it's done.
Inventor 2020, In-Cad, Simulation Mechanical
Just insert the picture rather than attaching it as a file
Did you find this reply helpful ? If so please use the Accept as Solution or Kudos button below.
Delta Tau Chi ฮฮคฮง
Unless I'm missing something spectacularly obvious, I'm gonna go out on a limb and say that I think Inventor has some potential optimisation issues. Check this out. Calling in @ChrisMitchell01 who might be interested in this, maybe bounce it to one of the team?
This bench score is from a Dell Precision T5810, it's brand new, I'm writing this post on it right now and it has:
Xeon E5-2687W @ 3.5GHz (10 cores, 20 logical cores)
Samsung NVME NAND SM951 M.2 PCIe SSD, rated and tested 2.1GB/s read and 1.5GB/s write speeds.
NVIDIA Quadro M4000
16GB DDR4 RAM @ 2133MHz
This is a ยฃ6000 workstation and by all measures is an absolute powerhouse of a system. After doing everything in my power to get the best out of it, testing registry tweaks, shutting all background services, multiple test cycles, this is the best it could do:
Now let's look again at my ยฃ350 naff in comparison Pentium build.
Pentium G3258 @ 4.2Ghz (2 cores, 2 threads)
Crucial BX200 SATA3 SSD, rated and tested at 540MB/s read and 490MB/s write speed.
AMD R9 380
8GB DDR3 1333MHz
Here's what it managed:
Ignoring the big IPI number, let's focus on some individual totals.
Drawing Test, how is it that when utilising a 10 core (20 threads) E5 V3 Xeon CPU, Inventor takes 25.48s to complete a MULTI-CORE/MULTI-THREADED PROCESS, a process which is widely accepted as being a function that uses multiple CPU cores. But my 2 core cheap Pentium finishes the exact same operation in 27.98s? I appreciate it isn't apples for apples, there's a lot of factors at play here, but seriously? Inventor has 20 threads to utilise and it can only execute a 100% CPU based operation 2 seconds faster than it could when it only had 2 threads? The Xeon is a ยฃ1800 product, the Pentium cost me ยฃ51.
How is it that when working on a SSD that is actually writing to the disk at 1566MB/s, Inventor took 3 seconds longer to save a file than it did when writing to a disk that only writes at 490MB/s and has significantly less IOPS values? This disk is writing 3 times faster than the other, yet Inventor took 3 seconds longer to write to that disk?
This is the Crystal Disk Mark report from the SSD in the Dell T5810.
Look I know this isn't apples for apples, I've said that. But everything in that Dell workstation is premium and professional, from the mobo to the PSU, to the chassis to the RAM, it's all premium grade high end stuff.
All jokes aside, how am I able to build a computer on a budget of ยฃ350 using low end consumer grade kit which is able to not only keep up with but out perform in some areas a workstation like this? It can't be the fault of the workstation, the power and grunt is actually there, is it that Inventor is just not optimised to use it?
I also know that the counter argument will be 'But that workstation will better handle large assemblies and the like'. Will it though? What if I put 16GB RAM in the cheap PC and made it a ยฃ380 PC, will it then handle large assemblies better? 'But it will compute complex FEA calculations faster'. Will it though? Because that's a 100% CPU intensive operation, just like creating drawing views is and look at how that turned out.
I'm honestly not beating on anyone or anything here, but I think some clarification would be nice as this is a total mystery to me at the moment. And given how many people have contributed to this thread, a lot of folk really care about this topic.
It took nearly an act of congress, but I actually got IT to install 2015 on my Lenovo P300 desktop (for "testing" purposes only), so I can finally run the benchmark tool.
Not very impressive results for the amount of money they spent on these machines.
And they expect us to consistently build 200,000 part assemblies with these. SAD 
I think part if the problem is back to the solid model, Inventor still needs to "build" the model with is the limited to a single core in the CPU before it can generate the drawing.
Possibly the model is closed, the majority of the time is used to build and assembly the model, once that is done, very little time is consumed in the drawing as it's spread across all the cores.
Inventor 2020, In-Cad, Simulation Mechanical
Just insert the picture rather than attaching it as a file
Did you find this reply helpful ? If so please use the Accept as Solution or Kudos button below.
Delta Tau Chi ฮฮคฮง
I don't think that's how this test works m8, the drawing test uses an already modelled part and the timer begins when the views are created placed. You can see the pre-modelled files in the bench folder in your documents area, VGAtest IPT's, it uses these to create the drawing views.
It's probably the worst time to be bringing this up though, we're right bang in the lead up to the 2017 product launch so it'll not be a priority for the team.
@mmaes The full specs are in the original post (a few pages back now) and in the description of the video I did, but the mobo is a mini ATX Asus H81M-PLUS. It's an entry level board, cheap and cheerful. Possibly one of the reasons why I couldn't OC the Pentium higher than 4.2GHz.
So I have a question about a coworkers high end laptop that he just got in. He built it with about every single high end part he could. He got it in and ran the benchmark test and couldn't get over a 5. I will list his specs and benchmark pic below. The differences between my desktop and his laptop are him having (2) GTX 980m GPU's vs my (1) GTX 980ti, and he has 64GB of Ram compared to my 32GB. His 6700k cpu is overclocked to 4.5ghz also. His scores are really low. They are just as low as his 3 year old Dell precision laptop. What could be wrong from the factory with this machine? Are the 980M GPU's significatly slower than the 980ti? Any help would be appreciated. Want to know if we need to send this back b/c it's not worth the $6000 we spent on it when it isn't running any faster than his old laptop.
Specs:
Builder - Origin
CPU - i7-6700K (overclocked to 4.5ghz)
GPU - (2) GTX 980m's
Ram - 64GB
Main HD - Samsung 950pro 512gb
Forums Links
Can't find what you're looking for? Ask the community or share your knowledge.
Post to forums
