Hello Dr. Smith,

I am considering purchasing a new workstation for Caedium.
As I understand it, mesh generation on Caedium uses but one core of even a multi-core computer, so a single (or fewer) core computer with faster processor core speeds is more helpful for mesh generation, whilst having many more cores is more helpful for solver running, even if the processor speeds are slower.

Is this correct? This is my main question, as it implies a need to balance number of cores and processing speed of each core.

I anticipate running future simulations up to, say, 5 million tetrahedra. How much memory is helpful to have? Running 2 million tetrahedra, it takes a long time to show the volume mesh on my computer with 16GB memory. It is so slow that I never save a Caedium simulation with the monitor showing the volume mesh, because it may be some time - sometimes maybe half an hour - before the computer actually shows it when I re-open up a simulation I have thus saved. I fully understand this is a reflection of my trying to do way too much with a laptop, rather than a problem with Caedium. I'm using a 4 core 2.8GHz MacBook Pro with 16GB memory. I have Caedium set to run on 3 of the 4 cores and 6 threads.

I'm curious, if it is not inappropriate to ask, what you use as your workstation for running Caedium? If that is proprietary, I apologize for asking.

Thanks very much for your help.