Marcus Eggenberger

Parallel Simulation of Networks on Chip


Front	Back

ISBN:

978-3-8440-7188-7

Series:

Informatik

Keywords:

NoC; Embedded Systems; High Performance Simulation; Shared Memory; Full System Simulation; Parallel Algorithms

Type of publication:

Thesis

Language:

English

Pages:

148 pages

Figures:

54 figures

Weight:

219 g

Format:

21 x 14,8 cm

Bindung:

Paperback

Price:

45,80 € / 57,30 SFr

Published:

February 2020

Buy:

DOI:

10.2370/9783844071887 (Online document)

Download:

Available PDF-Files for this title:

You need the Adobe Reader, to open the files. Here you get help and information, for the download.

These files are not printable.


	Document		Document
	Type		PDF
	Costs		34,35 EUR
	Action		Purchase in obligation and display of file - 2,2 MB (2297855 Byte)
	Action		Purchase in obligation and download of file - 2,2 MB (2297855 Byte)


	Document		Table of contents
	Type		PDF
	Costs		free
	Action		Display of file - 1,5 MB (1562938 Byte)
	Action		Download of file - 1,5 MB (1562938 Byte)

User settings for registered users

You can change your address here or download your paid documents again.

User:	Not logged in.
Actions:	Login / Register Forgotten your password?

Recommendation:

You want to recommend this title?

Review copy:

Here you can order a review copy.

Link:

You want to link this page? Click here.

Export citations:

Text
BibTex
RIS

Abstract:

Networks on Chip (NoCs) constitute a communication paradigm that scales both in performance and size and is thus suited for chips with 64 or more processor cores. Since NoCs offer a vast design space, fast simulation algorithms are an invaluable tool to ensure that design constraints are met and time to market is short. Although such many core systems have an inherent regular structure, simple parallel simulation algorithms fail to leverage the power of modern simulation machines. Imbalances in the simulated communication leads to imbalances in the workload for the simulator, which simple approaches cannot compensate.
This dissertation proposes a parallel simulation approach for Networks on Chips (NoCs) called Task-Based Simulation that achieves high simulation performance while maintaining cycle level accuracy. This is achieved by decomposing the simulation by space and time, which allows decoupling the simulation of consecutive cycles. Task-based simulation achieves speedups that are 1.5 to 2.4 times higher than those of the state of the art. And with overall speedups of up to 23.7x, task-based simulation provides cycle-accurate NoC simulation that is truly scalable.