The program offers several types of solvers. In general, they are categorized in two flavors: in-core and out-of-core solvers. With in-core solvers, the global building stiffness matrix is assembled, stored and solved in the physical memory (RAM) of the computer. As long as there is enough memory available for the solution of the models, this choice always gives the best performance\solution time. However, for very large models, the in-core solver might run into insufficient memory issues. If this is the case, it is suggested to switch to out-of-core solver. With the out-of-core solver, the program assembles stores and solves building stiffness matrix using files that are stored on hard-drive of the computer. Thus, it involves repeated access to hard drive, which may substantially increase analysis time. It is always recommended that models should be first run with the in-core solvers and if an out-of-memory error is detected, then the out-of-core should be used.

More specifically, the program includes direct and direct-sparse solvers. The first is suitable for moderate size problems where as the latter is for all types of models. Note that the sparse solvers are substantially faster solvers thus it is highly recommended for large models. Note that both types (direct and direct-sparse) include in-core and out-of-core versions.

If a sparse-solver type is selected, further options are provided: Use Single CPU Core or Use All Available CPU Cores. The former is intended for computers with old operating systems or with single core system or for a user who does not want to allocate all cores for the program’s use. The latter utilizes all available cores.

Note that no matter what solver type is used, they all lead to the same analysis results.