COMPUTING

PARADIGMS
DR. P. JEYANTHI
ASSISTANT PROFESSOR,
DEPARTMENT OF INFORMATION TECHNOLOGY
SRI RAMAKRISHMA COLLEGE OF ARTS & SCIENCE
Computing Paradigms
Computing Paradigms
• High-Performance Computing
• Parallel Computing
• Distributed Computing
• Cluster Computing
• Grid Computing
• Cloud Computing
• Biocomputing
• Mobile Computing
• Quantum Computing
• Optical Computing
• Nanocomputing
• Network Computing
 HIGH-PERFORMANCE COMPUTING
• Definition and Components:
• High-performance computing (HPC) involves a
network of processors (CPUs), memory, storage,
input/output devices, and software.
• It utilizes the entire connected system to execute
complex computational tasks.
• System Variety:
• HPC can include both homogeneous and
heterogeneous processor setups.
• Historically associated with supercomputers, HPC today
also encompasses smaller clusters, such as a group of
desktop PCs.
• Applications:
• HPC is commonly used in scientific research that
requires substantial computing power.
• Examples include simulating scientific phenomena, like
protein folding in molecular biology and nuclear
fusion modeling.
• Simulation Capability:
• HPC systems are especially suited for simulation
studies, enabling researchers to explore scientific
problems that are otherwise difficult to analyze.
 PARALLEL COMPUTING
• Definition:
• Parallel computing is an aspect of high-performance
computing (HPC) where multiple processors work
together to solve a computational problem.
• Processor Type:
• Typically, parallel computing involves homogeneous
processors, meaning the processors are similar or
identical in structure and performance.
• Relation to HPC:
• Parallel computing shares the same broad definition as
HPC, often including supercomputers with hundreds or
thousands of interconnected processors.
 DISTRIBUTED COMPUTING
• Definition:
• Distributed computing involves a network of multiple
computers or processors connected to function as a single,
unified system.
• Network Configuration:
• The connected systems can be either homogeneous (same type
of machines) or heterogeneous (different types, such as PCs,
workstations, and mainframes).
• Connectivity:
• CPUs in a distributed system can be physically close (e.g.,
connected by a local network) or geographically distant (e.g.,
connected via a wide area network).
• Goal:
• The primary objective is to make a collection of networked
machines work as one cohesive unit, creating a seamless and
efficient computing environment.
• Distributed computing systems are advantageous
over centralized systems, because there is a
support for the following characteristic features:
1. Scalability:
It is the ability of the system to be easily
expanded by adding more machines as
needed, and vice versa, without affecting the
existing setup.
2. Redundancy or replication:
Here, several machines can provide the same
services, so that even if one is unavailable (or
failed), work does not stop
 CLUSTER COMPUTING
• Definition:
• A cluster computing system is composed of multiple
similar or identical processor machines connected via
a dedicated network
• Resource Sharing:
• Nodes (individual computers) in a cluster share
common resources, such as a home directory, and
often use software like Message Passing Interface
(MPI) to allow simultaneous program execution across
all nodes.
• Type of HPC:
• Cluster computing is a type of high-performance computing
(HPC), as nodes work together to solve complex problems
that exceed the capabilities of a single computer.
• Node Communication:
• Nodes communicate with each other to coordinate and
cooperate, allowing them to solve larger problems
effectively.
• Heterogeneous Clusters:
• Clusters with heterogeneous processor types exist but
are mostly experimental or research-based.
 GRID COMPUTING
• Purpose:
• Grid computing aims to utilize underused computing
resources across organizations, increasing the return on
investment (ROI) by allowing organizations in need to access
this unused computing power.
• Definition:
• It involves a network of computing resources, or processor
machines, managed with specialized software (such as
middleware) to facilitate remote access and resource
sharing.
• Middleware & Grid Services:
• Middleware in grid computing is responsible for
managing resources, referred to as grid services.
• These services handle access control, security, data
access, and provide long-term storage solutions.
• Applications:
• Grid services can provide access to
data resources like digital libraries,
databases, and large-scale interactive
data for storage and processing needs.
• Grid computing is more popular due
to the following reasons:
• Its ability to make use of unused
computing power, and thus, it is a cost-
effective solution
• As a way to solve problems in line with any
HPC-based application
• Enables heterogeneous resources of
computers to work cooperatively and
collaboratively to solve a scientific
problem
 CLOUD COMPUTING
• Evolution from Grid Computing:
• Cloud computing evolved from grid computing, especially
when large-scale resources were needed to solve single,
complex problems.
• It builds on the idea of computing power as a utility.
• Difference from Grid Computing:
• While grid computing uses multiple computers in parallel to
address specific applications, cloud computing pools a
variety of resources (computing power, storage,
networks, etc.) to provide a unified service to end users.
• Dynamic Provisioning:
• Cloud computing can dynamically allocate IT and
business resources—such as servers, storage, and
networks—based on user needs and workload.
• Versatility:
• Clouds can provide both grid and nongrid services,
making them more versatile for different organizational
needs.
 BIOCOMPUTING
• Structuring Computer Programs:
• Biocomputing aids in developing computer programs and
applications based on biological models, which allows
computational simulations and problem-solving in ways that
mimic biological processes.
• Scientific Exploration:
• This field provides tools and theoretical foundations for scientists to
study and manipulate biological elements such as DNA and
proteins, which are fundamental to life.
• Focus on Order:
• Since the function of biological molecules depends heavily on the order
of their components, biocomputing researchers focus on determining
suitable molecular structures to mimic biological functions.
• Advancement in Disease Understanding:
• Through biocomputing, scientists gain insights into molecular
processes, which can improve the understanding of life and potentially
identify molecular causes of diseases.
 MOBILE COMPUTING
• Portable Computing Devices:
• In mobile computing, the processing elements are typically
small, handheld devices such as smartphones and tablets.
• Wireless Communication:
• Communication between devices and resources relies on
wireless media, allowing for connectivity and data exchange
without physical connections.
• Rapid Growth in Voice and Data Applications:
• Initially focused on voice communication (like cellular phones),
mobile computing has expanded significantly to support data
transmission, driven by the growing number of cellular network
subscribers.
• Data Transmission Capabilities:
• Mobile computing technology enables sending and receiving data
over cellular networks, which has led to applications such as
video calling, which extends beyond traditional voice-only calls.
• Remote Data Transmission:
• Mobile computing allows users to transmit data from
remote or mobile locations to other remote or fixed
locations, making it useful for a wide range of applications
that require mobility.
• Advancements and Applications:
• With continuous technological advancements, mobile
computing applications are evolving rapidly, offering new
functionalities and conveniences for users on the move.
 QUANTUM COMPUTING
• Limitations of Classical Computing:
• Traditional computing is reaching physical limits as
transistor sizes decrease, making it harder to continue
increasing processing power by doubling transistors
in integrated circuits.
• Quantum Computing as a Solution:
• Quantum computing offers a potential solution by utilizing
quantum information, based on the principles governing
subatomic particles, rather than conventional binary data
processing.
• Massive Speed Advantage:
• Quantum computers have the potential to be millions of
times faster than today's most powerful
supercomputers.
• Fundamental Differences:
• Quantum computing operates on fundamentally
different principles from classical computing,
which presents challenges for development and
adoption.
• Current State:
• While there are working prototypes of quantum
computers, they have not yet reached a stage
where they can replace traditional silicon-based
computers in practical, everyday applications.
 OPTICAL COMPUTING
• Use of Photons:
• Optical computing utilizes photons from visible light or
infrared (IR) beams for processing, rather than relying on
electric currents as in traditional electronic computing
• Speed Advantage:
• Since electric currents travel at about 10% of the speed of
light, optical computing promises much faster data exchange and
processing rates, potentially performing operations at speeds 10
times or more faster than conventional computers.
• Advantage of Optical Fibers:
• The limitations of electric currents in data transmission led to the
development of optical fiber technology, which optical computing
further leverages for faster and more efficient computation.
• Potential Impact:
• By implementing the benefits of visible light and IR networks at the
device level, optical computing could significantly increase
computational speed and efficiency.
 NANOCOMPUTING
• Nanoscale Components:
• Nanocomputing involves the use of components that
are only a few nanometers in size, often in at least
two dimensions.
• Carbon Nanotube Transistors:
• In nanocomputers, traditional silicon transistors may
be replaced by carbon nanotube-based
transistors, allowing for much smaller, efficient, and
potentially more powerful systems.
• Challenges in Scale and Integration:
• The success of nanocomputers depends on overcoming
issues related to the small scale of components and the
difficulty of integrating a large number of nanoscale
devices.
• Manufacturing Limitations:
• Economically feasible manufacturing of complex
patterns at such a tiny scale is challenging and
remains a significant obstacle
• Research and Development:
• Efforts are ongoing to address these challenges, as
researchers work on realizing nanocomputing systems
that could revolutionize the future of computing.
 NETWORK COMPUTING
• Leveraging Network Resources:
• Network computing designs systems to optimize the
use of advanced technology through a robust
network infrastructure, improving business solutions
and customer service.
• Client-Server Architecture:
• The client component resides with the end user,
providing essential functions at low cost and simplicity,
while the server environment delivers application
services to the client
• Minimal Configuration:
• Unlike traditional PCs, network computing clients
don't need individual configuration and
maintenance, making them easier to manage.
• Foundation of Modern Computing:
• Almost all modern computing paradigms are
networked, and future paradigms are likely to be based
on network architectures to maximize benefits for end
users.
• Broad Applicability:
• Network computing serves as a backbone for various
computing models, enhancing flexibility and scalability
across business applications and technology
deployments.