O'Reilly logo

High Performance Parallelism Pearls Volume One by James Jeffers, James Reinders

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 23

Characterization and Optimization Methodology Applied to Stencil Computations

Cedric Andreolli*; Philippe Thierry*; Leonardo Borges; Gregg Skinner; Chuck Yount    * Intel, France Intel, USA

Abstract

This chapter describes the characterization and optimization methodology applied to a 3D finite differences (3DFD) algorithm used to solve constant or variable density isotropic acoustic wave equation (Iso-3DFD).

Starting from the most basic implementation and to get close to its achievable performance we present a series of tuning steps from the basic version up to an implementation using intrinsic functions. The tuning techniques described here include scalable parallelization (collaborative thread blocking), maximizing memory bandwidth ...

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