a Motion stress-test …

… for a comparison of diff. hardware, mentioned here
www.fcp.co/forum/4-final-cut-pro-x-fcpx/...ac-vs-imac-pro#92568

… the idea came up for a 'Motion Stress Test', or as board member joema kindly phrased it,'KarstenX'

the first one in that other thread is probably too 'easy' for actual hardware.
so, here's 100cubes_vers2, a 10sec/30fps Motion project with some combo of nastiness :
100 individually particles of a 3D glyph, brushed Gold, affected by physics, lighted by 3 lights, cam sweeping around, and finally 15samples/180° MotionBlur…



after dl and unzip, Open the project-file in Motion5.4 and 'share it' (check, MotionBlur is set On) …

Post your hardware and results here …

2012 macMini, 2.5Ghz/16GB, intSSD = 2h 05min

… yes, two hours for 10secs of video is … not that impressive.
But some HAS to set the baseline.

Ok... I got 30m 10s on my Coffee Lake Hackintosh... (6-core, 3.7ghz, Radeon RX 580, 16gb ram)

But I kept the activity monitor open at the same time (and was browsing the internet as it's too long a test to sit and stare at it)... Motion was just averaging 150% CPU... that's just 1.5 out of 6 cores (it was using all 6 cores, just not pushing them). Why is Motion using so little of the available CPU power? Is it mostly the GPU working?

Yes, Motion utilizes the GPU a great deal now.

I'll be testing on my 2013 MP and in a couple of weeks on the new iMac Pro. Results to be published...

Looking forward to see results both for nMP and the iMac Pro!

Karsten: Would it be possible to make a test that isn't quite as long? Kinda like the BryceX that kan be done in under one minute on most newer machines? I don't think that many people will do the test if it takes up to an hour (or two, in your case)...

DmitriZigany wrote: …Karsten: Would it be possible to make a test that isn't quite as long? Kinda like the BryceX that kan be done in under one minute on most newer machines?…..

fully agree - and easy done: just switch-off Motion Blur

But no actual beast here at hand, so no clue how much is too much, or too little (my score in BruceX is 188secs LOL) - think, results ~2-3min should be fine, for differenciation … a minute seems to me too short, keep in mind, we check time manually …