Hi, Hyungsik -
If you're finding the GLPK solver slow, I suggest giving Cbc a try (which you can do by changing Transformation\Electricity Generation:Optimize to "NEMO(Cbc)"). In our experience, Cbc is often faster than GLPK for models with storage. When I solve the time slice demo POLSTORE scenario with Cbc, it takes 4-5 minutes (as opposed to 11 minutes with GLPK).
As a general matter, a faster computer is better for optimization modeling. However, for complicated models, the performance improvement from switching solvers is often greater than the improvement from upgrading hardware. LEAP and NEMO support several proprietary solvers (CPLEX, Gurobi, Mosek, and, coming soon, Xpress) that can offer orders-of-magnitude performance gains compared to Cbc and GLPK. If you plan on conducting a serious optimization modeling study, we suggest looking into acquiring one of these tools. If you're an academic user, note that you may be eligible for a free or discounted license. Check with the provider for licensing options.
Thanks,
Jason