[conformist]

Yes the combination of Krylov and quasi-Newton methods are very successful for physics problems (https://en.wikipedia.org/wiki/Quasi-Newton_method).

Iirc eg GMRES is a popular Krylov subspace method.