My guess is that grad QM is fairly similar at different schools. What I know (by comparing notes with my fellow first years) is that undergraduate preparation varies a great deal. People took between 1 and 3 QM classes, with different books, and teacher qualities. I'd say that repeat material ranged between 10% and 60% depending on the person.
As an undergrad, I took 2 10 week quantum courses. One was from Griffiths (isbn=0131118927)
, and the other was half from Sakurai (isbn=0201539292)
and half from Griffiths.
The grad QM class was mostly from Sakurai (although Shankar (isbn=0306447908)
was pretty useful. Sakurai will often comment on things which are completely obvious ("that's just the identity matrix divided by two." No ***.) but won't comment or provided examples on things which are very not obvious).
I'd say that there was maybe 50% overlap. Most of the stuff I'd seen from Sakurai, I saw again (mostly the formalism stuff like Dirac notation, and some other odds and ends like the path integral). However, there was less overlap with the Griffiths stuff -- Sakurai sticks all the wave function stuff in an apendix and expects you to know it already. So although we saw square wells, SHO, hydrogen atom, it was really building on what we knew already and there wasn't much overlap. The most direct overlap came when we did WKB. Most of the stuff in the last half of the grad class was all new to me (symmetries, spherical tensor operators, lots of angular momentum stuff (although a bit was in Griffiths), etc.)
I'll be taking grad QM II next term, and that may have more overlap with the second half of my second undergrad class (because QM II is more about applications/approximations).