Perfect!
After some testing, I have found that a buffer size of 48 is marginally faster than 64 and 32.
Is it automatically detecting the number of threads available?
Quite oddly, it seems that choosing different bucket orders has an impact on the rendering time.
This is rendered with column.
Hilbert : 39 s
Spiral : 39 s
Row : 38 s
Column : 35 s
Diagonal : 35 s
Random: 48 s
I didn't try the reverse options – but I must admit I was somewhat curious about the reverse random...
I noticed no difference with the threads.lowPriority activated.
The project file is meant as an illustration of the rather hot inner core of the earth (intended for the education of my daughters),
and I wanted it to spin, so I tried making a video.
It seems the video will not render in Q4, but stops about half way.
I'm including the project file, in case anyone wants to give it a try ( 1280 x 536 )
Project file
Cec
