Multiprocessors Cache Coherence Protocols-Parallel Processing-Lecture Slides, Slides of Parallel Computing and Programming

Prof. Bhairav Gupta delivered this lecture at Ankit Institute of Technology and Science for Parallel Processing course. It includes: Multiprocessors, Cache, Coherence, Protocols, Snoopy, Directory, Pentium, Invalidate, Processors

Typology: Slides

2011/2012

Uploaded on 07/23/2012

paramita
paramita 🇮🇳

4.6

(16)

120 documents

1 / 14

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Multiprocessors Cache
Coherence protocols
MESI PROTOCOL
Pentium processor
Snoopy
Write Invalidate
Write Back
Snoopy based protocols
Directory based protocols
docsity.com
pf3
pf4
pf5
pf8
pf9
pfa
pfd
pfe

Partial preview of the text

Download Multiprocessors Cache Coherence Protocols-Parallel Processing-Lecture Slides and more Slides Parallel Computing and Programming in PDF only on Docsity!

Multiprocessors Cache

Coherence protocols

MESI PROTOCOL •^

Pentium processor

•^

Snoopy

•^

Write Invalidate

•^

Write Back

•^

Snoopy based protocols

-^

Directory based protocols

Snoopy Cache Systems

How are invalidates sent to the right processors?In snoopy caches, there is a broadcast media that listens to all invalidates andread requests and performs appropriate coherence operations locally.

A simple snoopy bus based cache coherence system. Tags

Snoop H/W

Processor Cache

Tags

Snoop H/W

Processor Cache

Tags

Snoop H/W

Processor Cache

DirtyAddress/data

Memory

MESI protocol of Pentium Processor

MESI State Diagrams

•^

Initiating processor perspective

-^

Other processors perspective

Caused by other PEs cache controller snooping Activity on the bus

Read Hit

Write Miss

Example:

MESI protocol WorkingStep-by-step operations

Directory Based Systems

Architecture of typical directory based systems: (a) a centralized directory (b) a distributed directory.

(a)

(b)

Directory

Data

State

Presence

Bits

ProcessorCache ProcessorCache

ProcessorCache

ProcessorCache

Interconnection Network

Interconnection Network

Memory

Presence

bits / State

ProcessorCache Memory

Presence

bits / State

docsity.com

Performance of

Directory Based Schemes

-^

The need for a broadcast media is replaced by the directory.

-^

The additional bits to store the directory may add significantoverhead.

-^

The underlying network must be able to carry all the coherencerequests.

-^

The directory is a point of contention, therefore, distributeddirectory schemes must be used.

-^

Quantify the memory required for the Directory

-^

techniques for efficient directory implementation