GC is not a problem if you are repeatedly allocating large objects. The time to fill in the large objects massively dominates the GC time because GC effort is proportional mostly to the *number* of objects, not the total *size*.
That said, if you have large objects that vary in size you could be subject to heap fragmentation. That is a bit of an unusual problem. The rule of thumb is that people worry about GC cost far too much almost always. 2011/5/10 Sébastien Brisard <sebastien.bris...@m4x.org> > This allows reusing previously allocated vectors. I'm using these solvers > with large data sets (eg tomography), and I'm worried about the GC being too > heavily sollicitated if I keep allocating new vectors. For example, each > iteration of the iterative solver calls operate(x, y). >
