hi there,
I'm so loving PixInsight that i would dedicace a better processing rig to it
We've (for now) no GPU support in pixinsight, so all calculation is done on the CPU side.
Anyway, with AMD threadripper Gen3 CPU, there are now massive core count out there to get the job done.
And i know that there were hard job done by the PI core team to get better multithread support.
But as far as i know Amdahl's law draw some theorical limits in process parallelism return.
So here is my simple questions about a thread count :
Does it make senses to run pixinsight with a 64 ot 128 multithread architecture ? (AMD 3970 or 3990)
Is there any other software bottleneck that will hit us before that thread count ?