Parallel and distributed

computing
COMP3139
Agenda
• Basics of Data Communication

• Distributed computing use cases

• Mainframes

• Clusters

• Grid Computing

• Difference between Grid and Distributed

Computing

• Technologies for distributed computing

Basics of Computer Communication

• Many organizations use the internet and computer programs to talk to

each other and maintain communication over long distances.

• Understanding computer communication methods and how they function

may help you use them more effectively at work.

• Definition: Computer communications are any processes that allow you

to communicate which feature a computer or computer program. The
computer does this by sending data to one or more other computers, who
decode the message and allow the recipient to send a message back.
Importance of computer communications

• Using computer communication is important because it allows you to

send messages over long distances which can help you coordinate with
team members.

• Computer communication also usually has some form of built-in record

system .

• If you send an email to your supervisor about scheduling a meeting for

Monday, both you and your supervisor can reference that email later.
Computer communication types

Email

• Electronic mail, often shortened to email, sends a digital message to a

recipient in a process like traditional mail.

• You need to create a custom email address, which functions as a virtual

mailbox.

Text messaging

• Text messaging is like email, except you use a phone number instead of
an email address to determine where to send your message.
Computer communication types

Image and video messaging

• Image and video messaging are when you take a pre -taken picture or video
message and send it to another person.

Virtual reality chatting

• W ith new technology making virtual reality, also known as VR, accessible to
many people, many chat rooms allow a VR function

Livestream engagement

• Livestreams are a form of mixed computer communication where one person

or group streams content to a wider audience.
Distributed computing
use cases
• Distributed computing is everywhere
today
• Mobile and web applications are
examples of distributed computing
• Machines work together in the
backend for the application.
• Let’s explore some ways in which
different industries use high-
performing distributed applications.
Healthcare and life sciences
• Use distributed computing to:

o Model and

o simulate complex life science data

• Image analysis,

• Medical drug research, and

• Gene structure analysis all become faster with

distributed systems

• Accelerate structure-based drug design by

visualizing molecular models in three dimensions.
Healthcare and life sciences
• Reduce genomic data processing times to get early
insights into

o cancer, cystic fibrosis

o Develop intelligent systems that help doctors

diagnose patients by processing a large volume of
complex images like

o MRIs,

o X-rays, and

o CT scans.
Engineering research

• Engineers can simulate complex physics and mechanics concepts on distributed

systems

• To improve product design, build complex structures, and design faster vehicles

Here are some examples:

• Computation fluid dynamics research studies the behavior of liquids

• Implements those concepts in aircraft design and car racing.

Engineering research

Computer-aided engineering

• which requires compute -intensive simulation tools to test new

• plant engineering,

• electronics, and

• consumer goods
Financial services

• Firms use distributed systems to perform high -speed economic simulations

• assess portfolio risks,

• predict market movements, and

• support financial decision-making

• Create web applications that use the power of distributed systems

Financial services

Which can do:

• Deliver low-cost, personalized premiums

• Use distributed databases to securely support a very high volume of financial

transactions.

• Authenticate users and protect customers from fraud

Energy and Environment

• Need to analyze large volumes of data to improve operations and transition to

sustainable and climate-friendly solutions.

• Use distributed systems to analyze high -volume data streams from a vast
network of sensors and other intelligent devices

These are some tasks they might do:

• Streaming and consolidating data for the structural design of power plants

• Real-time oil well monitoring for proactive risk management

Mainframes
• Mainframes were the first example of large
computing facilities whic h leverage multiple processing
units

• Powerful, highly reliable computers specialized for

large data movement

• They are not considered as a distributed system;

however, they can perform big data processing

• Batch processing is the important application of

mainframes.
Clusters
• Clusters have started as the low-cost alternative
to the mainframes and supercomputer

• Since the 1980s cluster has become standard

technology for parallel and high-performance
computing.

• Due to their low investment cost different research

institutions, companies,

• cheap machines with high computational power

• Example of a cluster is weather forecasting

 Grid Computing
• Geographically distributed computer networks work
together to perform common tasks

• You can form them from computing resources that belong

to multiple individuals or organizations .

Grid computing can be used to:

• simulate the behavior of a nuclear explosion,

• model the human genome, or analyze

• data generated from particle accelerators.

 Grid computing vs. distributed computing

• Grid computing is highly scaled distributed computing that

emphasizes performance and coordination between several
networks.

• Each grid acts like a tightly coupled computing system,

Externally, grids are more loosely coupled

• Grid computing systems can be homogeneous or

heterogeneous and operate collectively as virtual
supercomputers .
 Grid computing vs. distributed computing

• Distributed computing systems consist of several

software components spread over multiple computers

• Distributed computing has a broad scope and can

describe systems in close physical proximity

• Distributed computing aims to achieve a single goal at

any one time.

• several distributed computing systems make a grid

computing system.
What is AWS High-Performance Computing?

• You can accelerate innovation with fast networking and virtually

unlimited distributed computing infrastructure.

For example:

• Amazon Elastic Cloud Compute EC2: to support virtually any

workload with secure, resizable compute capacity.

• AWS: to scale hundreds of thousands of computing jobs across

AWS compute services.
 Technologies for Distributed
Computing:
• Distributed computing, once a concept relegated to
the realm of research and specialized applications.

• Revolutionized data processing, storage, and

collaboration across industries and domains.

• Several key technologies driving distributed

computing
Technologies for Distributed Computing
Cloud Computing

• Cloud computing is perhaps the most prominent technology

underpinning distributed computing.

• Allows organizations and individuals to access and utilize

computing resources over the internet.

• Eliminating the need for on-premises infrastructure

• Cloud services such as Amazon Web Services (AWS)

Technologies for Distributed Computing
Cloud Computing

• Google Cloud provide scalable, on-demand

resources for everything from data storage to
running complex algorithms.

• Democratized access to high-performance

computing

• Enabling startups and large enterprises alike to

harness the power of distributed systems.

• Examples include Netflix, emails, MS office 365

THANK YOU