CS Elective 3 Week 1

Parallel Computing

simultaneously execution of multiple parts of a program using multiple processor or cores.

Parallel Computing

Divides The problem into smaller task executed at the same time instead of solving one problem step-by-step.

Image processing

if done through a parallel approach becomes significantly faster, While if done in a sequential approach one processor edits every pixel (slower).

Video Rendering

Movie Studio render animation frames simultaneously. instead of rendering frames 1 by 1 it becomes like this :

computer A- Renders 1-250

B- 251-500

C- 501-750

D- 751-1000

Rendering it like this makes it much faster (Parallel approach)

Grading 100 Exam papers

Sequential Approach : 1 teacher Checks all 100 papers (Consume too much time/ significantly slower)

parallel approach : The Papers are divided evenly among 4 teachers (Faster checking, does not consume too much time).

Early Computering (Single system Processor)

Early Computers use only one processor (CPU) to execute instructions. Each program are processed one at a time, making execution for complex computations relatively slow.

Characteristics of Single processor System

*Single CPU

*Sequential Execution

*Limited Multitasking

*Low Computational Power

Example of Early Computing

-A Calculator: Performs one arithmetic operation after another, Each operations waits for the previous one to finish.

Multi-Core Computing

-Contains Multiple processor cores inside a one CPU, each core can execute instructions simultaneously.

Example of Multi-Core computing

-A Quad Core processor

core 1: Play musiccore

2: Download Filecore

3: Edit DocumentCore

4: Antivirus scan

All task are executed concurrently instead of one at a time.

High Performance Computing (HPC)

-Combines many processors to solve Extremely complex problems.

Application of HPC

Climate prediction

Space exploration

Drug discovery

Artificial intelligence

Example of HPC

NASA uses super computers to simulate rocket launches before Actual testing.

Distributed Computing

Uses multiple independent computers connected through a network to solve one problem.

Example of Distributed computing

-Google Searches

-Facebook

-Online Banking

Advantages of parallel computing

-Faster Execution

-better resource Utilization

-Handles large problems

Faster Executions

Rendering Movies takes days on a single processor but only takes hours using thousand of processors.

Better Resource Utilization

Multiple CPU cores are used instead of leaving some Idle

Handles Large Problems

Examples:

Artificial intelligence

Machine Learning

Weather Forecasting

DNA Sequencing

Advantages of Distributed Computing

-Resource Sharing

-Fault Tolerance

-Scalability

Resource sharing

Computers Share storage and processing power

Ex. Google Drive.

Fault Tolerance

If one server Fails Another can continue the service

ex. Netflix

Scalability

New Servers can be added when demand increases.

ex. Companies add more cloud servers during online shopping events to handle Increased traffic.