Is there any available information regarding efficiency of Dmol3 and Dmol3-Solid parallelization? I want to know how many processors should I use for maximum efficiency.