I'm thinking of upgrading my machine for faster rendering and have been looking at the resources APG uses with Windows Task Manager/Resource Monitor. It seems that during the initial optimisation phase APG only uses the CPU and RAM (i.e. no disks). What puzzles me is why CPU utilisation is only around 12%? All 8 cores are being used but not very much. What is the bottleneck here? My intuition says that if optimisation is inherently sequential in a SISD sense then there can only be one thread (if I am using the term correctly). So only 1 out of 8 cores will be used i.e. 12.5%. And the fact that all 8 cores are being used is just a side-effect of the machine's system programming?
Is this correct? Does it even make any sense? I have a Ph.D in theoretical computer science but my knowledge of how computers work is stuck in the days of the Z80 :-)
If 'yes', am I right in thinking that as far as optimisation is concerned, the only way to get speed up is to have a CPU with good straight-line speed? This seems fairly obvious but as I say: I don't know such things - I only do eyes.
Of course optimisation is only one (small) aspect of what APG does and I know that blending is where I need to focus my attention if I want to make significant improvements to overall run-time. But I am interested to know if I am starting to understand how modern computers work.