These are papers I wrote or contributed to that discuss a Sandia-centric view of the state of parallel computing. I freely admit to a strong distributed-memory message-passing bias. That is, I've never thought the following concepts were that useful in practice: shared-memory machines, SIMD programming methodologies, or automatic parallelizing compilers.
Of course when you work at Sandia with a series of big distributed-memory machines and everyone lives and breathes message passing, you tend to look at the world as one big nail to pound your favorite hammer on!
Massively Parallel Computing: A Sandia Perspective, D. E. Womble, S. S. Dosanjh, B. A. Hendrickson, M. A. Heroux, S. J. Plimpton, J. L. Tomkins, D. S. Greenberg, Parallel Computing, 25, 1853-1876 (1999). (abstract)
Massively-Parallel Methods for Engineering and Science Problems, W. J. Camp, S. J. Plimpton, B. A. Hendrickson, R. W. Leland, Communications of the ACM, 37, 31-41 (1994). (abstract)
Is SIMD Enough for Scientific and Engineering Applications on Massively Parallel Computers ?, S. J. Plimpton, S. S. Dosanjh, R. B. Krall, in Proc of 37th IEEE International Computer Conference, COMPCON '92, San Francisco, CA, February 1992. p 95-102. (abstract)