I assume that my material has ferromagnetic properties. So, during it´s geometry optimization in CASTEP calculation, I chose to check the spin polarization option. But after about 3 hours the job failed and it´s failing each time with spin being on. However, the job is running smoothly if keep the spin polarization off and giving successful outputs very fast. I was working with a metal cubic structure with k points 6 6 6 and was trying to figure out it´s cut off energy through convergence test.