Parallel and Distributed Computing

CS 3006 (BCS-7A | BDS-7A)

Lecture 1
Danyal Farhat
FAST School of Computing
NUCES Lahore
Introduction to Parallel and
Distributed Computing
Motivation for Parallel and Distributed Computing
• Uniprocessor are fast but
Some problems require too much computation
Some problems use too much data
Some problems have too many parameters to explore
• For example
Weather simulations, gaming, web Servers, code breaking
Need for Parallelism?
• Developing parallel hardware and software has traditionally been
time and effort intensive
• If one is to view this in the context of rapidly improving
uniprocessor speeds, one is tempted to question the need for
parallel computing
• Latest trends in hardware design indicate that uni-processors may
not be able to sustain the rate of realizable performance
increments in the future
• A number of fundamental physical and computational limitations
• The emergence of standardized parallel programming
environments, libraries, and hardware have significantly reduced
time to develop (parallel) solution
Moore’s Law
• Number of transistors incorporated in a chip will approximately
double every 18 months.
Moore’s Law (Cont.)
Doubling the transistors does doubles the speed?
• True for single core chips but now the increase in number of
transistor per processor is due to multi-core CPU’s
• In multi-core CPUs, speed will depend on the proper utilization
of these parallel resources
Will Moore’s law hold forever?
• Adding multiple cores on single chip causes heat issues
• Increasing the number of cores, may not be able to increase
speeds [Due to inter-process interactions]
• Transistors would eventually reach the limits of miniaturization
at atomic levels
Need for Parallel and Distributed Computing
• Need for:
Good parallel hardware
Good system software
Good Sequential algorithms
Good Parallel algorithms
• Parallel and Distributed Systems
Parallel and Distributed Systems
• Parallel and distributed computing is going to be more and
more important
Dual and quad core processors are very common
Up to six and eight cores for each CPU
Multithreading is growing
• Hardware structure or architecture is important for
understanding how much it is possible to speed up beyond a
single CPU
• Also capability of compilers to generate efficient code is very
important (e.g. converting for loop into threads)
• It is always difficult to distinguish between HW and SW
influences
Distributed System vs Distributed Computing
Distributed Systems:
• “A collection of autonomous computers, connected
through a network and distribution middleware.”
• Enables computers to coordinate their activities and to
share the resources of the system
• The system is usually perceived as a single, integrated
computing facility
• Mostly concerned with the hardware-based accelerations
Distributed System vs Distributed Computing (Cont)
Distributed Computing:
• “A specific use of distributed systems to split a large and
complex processing into subparts and execute them in
parallel to increase the productivity.”
• Computing mainly concerned with software-based
accelerations (i.e., designing and implementing algorithms)
Parallel vs Distributed Computing
Parallel Computing:
• The term is usually used for developing concurrent
solutions for following two types of the systems:
Multi-core Architecture
Many core architectures (i.e., GPU’s)
Distributed Computing:
• The type of computing mainly concerned with developing
algorithms for the distributed cluster systems
• Here distributed means a geographical distance between
the computers without any shared-Memory
Parallel vs Distributed Computing (Cont.)
Parallel Computing: Distributed Computing:
• Shared memory • Distributed memory
• Multiprocessors • Clusters and networks of
• Processors share logical workstations
address spaces • Processors do not share
• Processors share physical address space
No sharing of physical memory
memory
• Also referred to the study • Study of theoretical
of parallel algorithms distributed algorithms or
neural networks
Practical Applications of P&D Computing
Scientific Applications:
• Functional and structural characterization of genes and
proteins

• Applications in astrophysics have explored the evolution of

galaxies, thermonuclear processes, and the analysis of
extremely large datasets from telescope
Practical Applications of P&D Computing (Cont.)
Scientific Applications:
• Advances in computational physics and chemistry have
explored new materials, understanding of chemical
pathways, and more efficient processes
Example: Large Hydron Collider (LHC) at European Organization for Nuclear
Research (CERN) generates petabytes of data for a single collision
• Weather modeling for simulating the track of natural
hazards like extreme cyclones (storms), flood prediction
Practical Applications of P&D Computing (Cont.)
Commercial Applications:
• Some of the largest parallel computers power the Wall
Street!
• Data mining-analysis for optimizing business and
marketing decisions
• Large scale servers (mail and web servers) are often
implemented using parallel platforms
• Applications such as information retrieval and search are
typically powered by large clusters
Practical Applications of P&D Computing (Cont.)
Computer Systems Applications:
• Network intrusion detection: a large amount of data needs
to be analyzed and processed
• Cryptography (the art of writing or solving codes) employs
parallel infrastructures and algorithms to solve complex
codes – SHA 256
• Graphics processing – Each pixel represents a binary
• Embedded systems increasingly rely on distributed control
algorithms - Modern automobiles use embedded systems
Thank You!