Please use this identifier to cite or link to this item: http://dx.doi.org/10.14279/depositonce-6784
Main Title: Extending the Cell SPE with Energy Efficient Branch Prediction
Author(s): Briejer, Martijn
Meenderinck, Cor
Juurlink, Ben
Type: Book Part
Language Code: en
Is Part Of: 10.1007/978-3-642-15277-1
Abstract: Energy-efficient dynamic branch predictors are proposed for the Cell SPE, which normally depends on compiler-inserted hint instructions to predict branches. All designed schemes use a Branch Target Buffer (BTB) to store the branch target address and the prediction, which is computed using a bimodal counter. One prediction scheme pre-decodes instructions when they are fetched from the local store and accesses the BTB only for branch instructions, thereby saving power compared to conventional dynamic predictors that access the BTB for every instruction. In addition, several ways to leverage the existing hint instructions for the dynamic branch predictor are studied. We also introduce branch warning instructions which initiate branch prediction before the actual branch instruction is fetched. They allow fetching the instructions starting at the branch target and thus completely remove the branch penalty for correctly predicted branches. For a 256-entry BTB, a speedup of up to 18.8% is achieved. The power consumption of the branch prediction schemes is estimated at 1% or less of the total power dissipation of the SPE and the average energy-delay product is reduced by up to 6.2%.
URI: https://depositonce.tu-berlin.de//handle/11303/7570
http://dx.doi.org/10.14279/depositonce-6784
Issue Date: 2010
Date Available: 12-Apr-2018
DDC Class: 004 Datenverarbeitung; Informatik
Subject(s): branch prediction
branch instruction
cell processor
branch predictor
branch target
License: http://rightsstatements.org/vocab/InC/1.0/
Book Title: Euro-Par 2010 - Parallel Processing
Editor: D’Ambra, Pasqua
Guarracino, Mario
Talia, Domenico
Publisher: Springer
Publisher Place: Berlin; Heidelberg
Publisher DOI: 10.1007/978-3-642-15277-1_29
Page Start: 304
Page End: 315
Series: Lecture Notes in Computer Science
Series Number: 6271
EISSN: 1611-3349
ISBN: 978-3-642-15276-4
978-3-642-15277-1
ISSN: 0302-9743
Appears in Collections:FG Architektur eingebetteter Systeme » Publications

Files in This Item:
File Description SizeFormat 
10.1007.978-3-642-15277-1_29.pdf708.03 kBAdobe PDFThumbnail
View/Open


Items in DepositOnce are protected by copyright, with all rights reserved, unless otherwise indicated.