For a real answer, you need to be more specific. We are providing for openLCA databases that are, as such, not available for other LCA software because they are really "heavy", such as PSILCA or full exiobase. If you run these on a computer that is quite weak, then this does not really work. Also, you need to allocate the memory that the systems you model and calculate need. You can, even for a not-powerful computer, use the big databases, but model with a cut-off.
1) not enough memoy available or allocated?
2) every new calculation also needs some memory, thus as in 1)
3) product system creation is not necessarily a simple step. In PSILCA developer, you have 100 million connections in a system.