The middle third of the lectures on Parallel Computing is concerned with efficient, practical use of real, parallel computing systems with a shared memory through which processor(-core)s can exchange information. It deals first with structure and properties of and inevitable constraints on shared memory systems that must be understood and taken into account in the development of algorithms and implementations for such systems and distinguishes real shared memory systems from idealized constructs like the PRAM. As concrete programming frameworks for such systems that illustrate many fundamental ideas, the library pthreads and the programming language extension OpenMP (and, briefly touched upon, Cilk) are treated in detail.