Vectorization (parallel Computing) - General Framework

General Framework

The general framework for loop vectorization is split into four stages:

  • Prelude: Where the loop-independent variables are prepared to be used inside the loop. This normally involves moving them to vector registers with specific patterns that will be used in vector instructions. This is also the place to insert the run-time dependence check. If the check decides vectorization is not possible, branch to Cleanup.
  • Loop(s): All vectorizes (or not) loops, separated by SCCs clusters in order of appearance in the original code.
  • Postlude: Return all loop-independent variables, inductions and reductions.
  • Cleanup: Implement plain (non-vectorized) loops for iterations at the end of a loop that are not a multiple of the vector size) or for when run-time checks prohibit vector processing.

Read more about this topic:  Vectorization (parallel Computing)

Famous quotes containing the words general and/or framework:

    Surely one of the peculiar habits of circumstances is the way they follow, in their eternal recurrence, a single course. If an event happens once in a life, it may be depended upon to repeat later its general design.
    Ellen Glasgow (1873–1945)

    Seeing our common-sense conceptual framework for mental phenomena as a theory brings a simple and unifying organization to most of the major topics in the philosophy of mind.
    Paul M. Churchland (b. 1942)