Chapter 2

Device architectures

Abstract

This chapter describes various architectural techniques for exploiting parallelism. It then provides an overview of processor designs in various market segments and identifies trends in design choices.

Keywords

Architectures

Cache

design

Memory

Register

SIMD

SMT

Thread

VLIW

2.1 Introduction

OpenCL has been developed by a wide range of industry groups to satisfy the need to standardize programming models that can achieve high performance across the range of devices available on the market. Each of these companies has specific goals in mind for OpenCL and targets for what features OpenCL should have to be able to run correctly on a specific architecture. To this end, OpenCL has a range of features that ...

Get Heterogeneous Computing with OpenCL 2.0 now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.