Multiphysics-based　optimization　is　a　useful　technique　for　high　performance　microwave　design.　An　overview　for　recent　advancements　in　multiphysics　design　using　space　mapping　(SM)　techniques　is　provided　in　this　paper.　We　develop　a　surrogate　model　including　two　mapping　modules　and　an　EM　coarse　model　using　SM.　This　technique　provide　a　effective　and　larger　update　for　every　iteration　by　utilizing　the　information　of　the　coarse　model　represented　using　inexpensive　EM　domain　samples,　resulting　in　optimal　solutions　being　obtained　more　efficient　than　the　existing　methods.　A　high　power　microwave　filter　is　used　to　illustrate　this　technique.　©　2021　Applied　Computational　Electromagnetics　Society　(ACES).