Quote - Okay... I don't know if this is correct or not; but, on the issue of the efficiency stat; according to Lux's website, the higher your efficiency is the easier it is for Lux to find and process your lights.  Since finding those lights seems to be about 90% of the game with Lux this translates to your render looking like a render with fewer passes.  I did some testing last night and made an excel spreadsheet with the details.  I was rendering the daz3d '61 vette and made 5 minute test renders after making a string of changes to the materials on the car itself.  Then rendered with and w/o gpu accelleration and with and w/o rendering in a cube.  You can see the massive increase in efficiency when rendering in a cube.   Sadly all the renders were done with the same file name so no screenies to show the results.  BUT, without the cube and with low efficience; I was getting more samples/sec and more passes; but after 5 minutes my car was still very amorphous.  With the cube and with massive efficiency, my car was looking like a car after just a couple passes.

 

Here a LINK to the spreadsheet on my google docs for those who may be curious enough to check it out.

 

Thanks for the spreadsheet sharky! I do so wish you had the renders though.

I must say, your data has made me curious of the efficency versus speed argument I've debated with myself. I've pounded it into my head that the ks/s was the key, and much to my surprise, if may not be the golden egg I thought it was. Hmm!

 

While reading through posts, I don't recall who mentioned it, but I'm also using an AMD 1090T Black. It handles overclocking quite well, but I'm watercooled too. Makes a difference this time of year here in Texas!

crimsonworx.com; free ebooks and previews

I've bowed down to facebook: https://www.facebook.com/crimsonworx