Gene Amdahl, "Validity of the Single Processor Approach to Achieving Large-Scale Computing Capabilities", AFIPS Conference Proceedings, (30), pp. 483-485, 1967
Next, because I want to go over something easy, I went to look at lock-free algorithms. i have heard of this once before in the context of kernel operations, and I always wanted to know how it worked.
http://en.wikipedia.org/wiki/Lock-free_and_wait-free_algorithms
contains a simple description for lock-free algorithms, along with some references on how they are designed. Abandoning this search for now, though it might prove useful for something like ParFUM?
Might be VERY useful, contains references to transactional memory and how it's used in concurent systems. Samoe comcurrent versions of programs already have constructs for this.
lots of references to languages and bibliographies. Good reading material