Well, this has been interesting.
I've got 3 Inventor systems, and I've played around with some hardware configurations for them, settings, etc. The results have been very illuminating. Numbers below are what I got from the spreadsheet output from Inventor Bench on all 3 systems.
System 1: Toshiba Satellite P755-S5269. i7-2630QM / GT540M / 8GB RAM / Samsung 840 250 GB(original, not evo) / 750 GB platter drive instead of DVD burner. Updated to Windows 10 Home, but last clean install was 2012 when I got the SSD. Predictably, this isn't a rocketship compared to my home office desktop, but Inventor is perfectly usable, particularly with smaller models and assemblies and there were a few major surprises with it. This is what runs my CNC router at home with Mach3, and I do a fair amount of Inventor work on it in the garage, particularly generating and tweaking toolpaths with HSM.
System 2: Home Office Desktop. i7-4770k / Radeon R9-270X / 16GB / Samsung 850 EVO 500GB / Samsung 840 250 GB (not evo) / misc platter drives / ASUS Z87-Pro motherboard / 2 monitors at 1080P, 1 at 1600x1200. Windows 10 Home, upgraded. The numbers shown here are from a test using the 840 as my system drive, with 66.9 GB free space remaining. Once the test was concluded, I used the Samsung drive migration software to clone that drive over to my new 500 GB 850 EVO, and re-ran the test. Unfortunately for some reason I didn't save the data, but I was surprised to see that the numbers were in fact just a hair lower on the brand new drive with 300+ GB of free space. I'll be re-running this test over the weekend, possibly in both drive configurations, with the addition of another 16 GB of RAM.
System 3: Day Job Workstation. HP Z220 workstation. Xeon E3-1240v2. AMD FirePro V3900. 1TB platter drive. Factory memory configuration was 2 sticks of 2GB each, 1 of 4GB. I ran the test both with those 8 GB and again with the 4GB removed and replaced with 2x8 of DDR3-1600. This is running Windows 10 Pro, upgraded from Win 7 Pro. This test had some pretty significant surprises, and made me re-evaluate some of the results from my other systems as well.
So. What were the surprises?
- I was surprised to find that my office workstation performed significantly better in almost everything except HD access time with only 8GB instead of 20. That really blew my mind. The difference is marginal, but it's still a difference.
- Graphics has a huge weight on how Inventor Bench calculates your Inventor Performance Index. If you compare the performance between my laptop and my office workstation, the laptop gets destroyed in every single category ... except for total graphics time. The margins aren't even close. Most of the time it's a difference of 50% or more ... but the IPI score for the laptop is actually higher than the score for the workstation: 3.89 vs 3.79/3.75.
- Speaking of graphics, good god - how does an entry level laptop graphics card smash a workstation card that badly, performance-wise? I'd expected the R9-270X to beat up on the FireGL V3900 pretty solidly, but I did NOT expect the laptop to win the same matchup.
- Model Save Time / Total HDD time was far worse than I'd expected on the two SSDs, and far better than I'd expected on the platter drive in my workstation. It's nothing special, just a 7200 RPM WD Blue 1TB. I haven't defragged it in, I dunno, probably a year or more now. Somehow it beat the SSD in my home desktop by well over a second, and it came in at almost HALF the time of the SSD in my laptop. (Which, BTW, isn't even close to being full.)
I'm looking forward to the results of the next batch of testing over the weekend. We'll see what the differences - if any - wind up being.
Rusty
) I have a 4790K and a GTX970 at home which I would love to test but I dont have Inventor on that computer sadly.
