KeyShot's Benchmark Testing Results

HaroldL · October 24, 2018, 09:09:05 AM

I noticed on the KeyShot Benchmark Testing doc located in the Media section of the Downloads page that there is a change in the results from FPS to Time. With the big hype for FPS in the Benchmark section of the forum, can someone explain the reason for the change?

Rex · October 24, 2018, 12:05:23 PM

Hi Harold,

The FPS values are on the back of the handout you've linked to. We are currently working on re-doing/updating this document as it was a poorly executed delivery by a previous member of the Luxion team.

HaroldL · October 24, 2018, 01:51:13 PM

Thanks for the feedback Rex. Do you have an idea when the doc will be updated? At my work we started looking at it to help specify new workstations without breaking the budget.

Rex · October 24, 2018, 05:00:54 PM

Actually it would be great to get your feedback to help push it along.

Are there any constraints on your purchase decision aside from price? i.e. Are you limited to certain manufacturers? Are you looking for a workstation, small form factor, or laptop? Windows or Mac?
Is this a dedicated KeyShot rendering machine or will it be used for other software? Do you have GPU requirements?

Thanks!

joseph · October 24, 2018, 10:54:03 PM

Hi Rex, is the turbine model bip still available? regards

mattjgerard · October 25, 2018, 05:41:28 AM

Quote from: Rex on October 24, 2018, 05:00:54 PM
Actually it would be great to get your feedback to help push it along.

Are there any constraints on your purchase decision aside from price? i.e. Are you limited to certain manufacturers? Are you looking for a workstation, small form factor, or laptop? Windows or Mac?
Is this a dedicated KeyShot rendering machine or will it be used for other software? Do you have GPU requirements?

Thanks!

Corporate IT loves HP, so I'm fairly limited

FUll blown under desk foot warmer workstation. Don't care what it looks like or how big it is. Noise not a huge factor, but as quiet as reasonably possible

Winders, unless the macs can beat the core count/speed (which the can't, so, windows) and corporate IT frowns upon anything without a 5 year warranty so no hackintoshes

70% KS, 20% Cinema 4d (lots of single threaded operations) 10% photoshop

No real GPU requirements other than what KS needs.

theAVator · October 25, 2018, 09:33:12 AM

I guess for me, both FPS and Time would be good to show. I think our IT guys/gals would be more adept at seeing the difference in times and better see how that could affect our productivity - or what a real-world tangible outcome is.

Our IT is pretty much set with Dell. There's a bit of an initiative corporate-wide to purge all of the old HP's we have and replace with newer Dells.
Right now our standard engineering workstation is a Dell T5810 with 3.0 GHz Xeon E5-1660 v3 and 64 GB RAM, NVIDIA Quadro M4000. No slouch, but when it has to be your computer for everything you do, any reduction in render times is an improvement in overall use. But our IT also hasn't really had to deal with CPU based renderers that can utilize 100% CPU either, so that document is a great help in visualizing the differences.

Rex · October 25, 2018, 10:05:38 AM

We do not have rights to the turbine model so cannot share it. Sorry. We will make sure all scenes included in the new tests can be shared freely.

That being said, what are the types of scenes you all would like to see tested?

This is all great feedback and much appreciated!

mattjgerard · October 25, 2018, 12:53:01 PM

Simple Diffuse
Metallic With roughness and bump maps
Cloudy plastic with lights behind
Displacement
Scattering with Spotlight

Or have one scene with different model sets that add complications to it so we can see what really slows things down.