Latest versioin of Arnold C4DtoA 3.3.4..Strange rendering speeds. An identical vdb frame renders on an i7 6 core at 3.5 Ghz in 1m 18 secs and 1m 46 secs on a Dual Xeon E-2670 - 16 cores at 3.00 GHZ The Xeon is in fact reading from an SSD so disk load cannot be a bottleneck as the I7 reads from a standard SATA drive. The Xeons have 128 GB RAM the I7 64 GB RAM... the I7 RAM is slightly slower but is running at 2T whereas the Xeon RAM is at 1TIs this an issue with dual CPU's and Arnold/Cinema 4dIn general on normal renders the dual Xeon renders twice as fast as the I7
The first thing to do is get detailed Arnold logs for two render.
As mentioned earlier the issue only seems to be with volumes. Otherwise the Xeon renders twice as fast as the I7
Is must be a vdb Arnold volume issue as rendering Turbulence FD containers direct using bcf file I get the speed up from the dual xeon
Thanks for the logs. They show that the Xeon is doing significantly more work (tracing more shadow volume rays, doing more volume marching samples, and more volume shader calls), so it's no surprise that it's rendering more slowly despite being the more power machine.
Why this is happening I don't know. Could it be something got modified when the scene was copied from one computer to another? Do the rendered images look identical?
So we can remove more unknowns, can you export to .ass file and render that with kick? You'll want to copy the .ass file from one computer to the other so that we're sure it's the same scene being rendered on both machines.
And there are different versions of CINEMA 4D on each machine.
i7 has R23, dual Xeon has R12