O'Reilly logo

Automatic Parallelization by Samuel P. Midkiff

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

CHAPTER 4

Transformations to modify and eliminate dependences

If an original, unchanged loop is analyzed, dependences will almost certainly be found that prevent useful transformations, such as tiling, interchange and parallelization, from being performed. In this chapter, several transformations that modify, or completely eliminate dependences will be described. These transformations typically do not lead to an increase in performance—on the contrary, they may lead to more loop bounds and subscript index expressions and computation, resulting in more work being performed. They can, however, enable transformations that dramatically increases the performance of the program.

There is another class of transformations—reduction and recurrence recognition—that ...

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required