The Lorenz attractor is the archetypal example in Julia documentation for solving systems of ODEs using DifferentialEquations.jl. Suppose I would like to solve that system for one thousand different initial conditions, keeping the attractor parameters fixed. What would be a good (i.e. efficient) way to solve and store the data for these 1000 decoupled ODEs?
Answering my own question...
A very interesting option is to work with ensembles:
https://diffeq.sciml.ai/stable/features/ensemble/
This feature offers support for various data analysis as well as good performance by using multi thread functionality. Should be a good starting point for anyone interested in running what in the original question should be called an Ensemble.