Module 1-Part 1

download Module 1-Part 1

of 16

Transcript of Module 1-Part 1

  • 8/3/2019 Module 1-Part 1

    1/16

    Mr. S.S. Sayyad

    Assistant Professor, ADCET Ashta

    [email protected]

    Module 1

    Introduction to Parallel Processing

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    2/16

    Trends towards Parallel Processing

    Data Processing

    Information

    Processing

    Knowledge

    Processing

    Intelligence

    Processing

    IncreasingVolumesofRaw

    Materialtobeprocessed

    IncreasingComple

    xityand

    sophisticationinprocessing

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    3/16

    Trends Towards Parallel Processing

    Data processing: data space is largest, including numeric numberin various format, character symbols, and multidimensional

    measure

    Information processing: Collection of data objects that are

    related by some syntactic structure or relation.

    Knowledge processing: consist of information items plus somesemantic meanings.

    Intelligence processing: is derived from a collection of

    knowledge items.

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    4/16

    Trends towards Parallel Processing

    From Operating Systems point of view, computer systemsare improved in four phases

    Batch Programming

    Multiprogramming

    Time Sharing

    Multiprocessing

    Parallel Processing is an efficient form of information

    processing which emphasizes the exploitation of concurrent

    events in the computing process.

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    5/16

    Trends towards Parallel Processing

    Parallel programming can be challenged in fourprogrammable levels

    Job or Program Level

    Task or Procedure Level

    Inter instruction Level

    Intra instruction Level

    Hardware Software Trade Offs

    Parallel and Distributed processing

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    6/16

    Parallel Computer Architectures

    Pipeline Computers: A pipeline computer performsoverlapped computations to exploit temporal parallelism.

    Array Processors: An array processor uses multiple

    synchronized arithmetic logic units to achieve spatial

    parallelism

    Multiprocessor Systems: A multiprocessor system achieves

    asynchronous parallelism through a set of interactive

    processors with shared resources.

    Data Flow Computers

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    7/16

    Department of CSE, ADCET, Ashta

    Pipeline

    Computer

  • 8/3/2019 Module 1-Part 1

    8/16

    Department of CSE, ADCET, Ashta

    Array

    Processor

  • 8/3/2019 Module 1-Part 1

    9/16

    Multiprocessor Systems

    Department of CSE, ADCET, Ashta

    Aimed at : Improving throughput, reliability, flexibility andavailability.

    Basic multiprocessor system contain two or more processor.

    All processor share access to common sets of memory

    module, I/O channels and peripheral devices. Most importantly, whole system is controlled by a single

    integrated operating system.

    Beside the shared memories and I/O devices, each processor

    has its own local memory and private devices. Inter processor communication can be done throw shared

    memory or throw interrupt network.

  • 8/3/2019 Module 1-Part 1

    10/16

    Data Flow Computers

    Programs for Data Driven Computation

    Calculation of z= (x+y)*2

    Template Implementation

    Basic Data Flow Mechanism

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    11/16

    Multiplicity of Instruction Data Streams

    Single Instruction Stream- Single Data Stream

    Processing Units

    Control Units

    Memory Modules

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    12/16

    Multiplicity of Instruction Data Streams

    Single Instruction Stream- Multiple Data Stream

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    13/16

    Multiplicity of Instruction Data Streams

    Multiple Instruction Stream- Single Data Stream

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    14/16

    Multiplicity of Instruction Data Streams

    Multiple Instruction Stream- Multiple Data Stream

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    15/16

    Serial v/s Parallel ProcessingFengs

    Scheme

    Use degree of parallelism to classify various computerarchitectures.

    Maximum number of binary digits (bits) that can beprocessed within a unit by a computer system is called the

    maximum parallelism degree P. Average Parallelism Degree Equation:

    Processing Methods and their Examples (text)

    Word Serial Bit Serial

    Word Parallel Bit Serial Word Serial Bit Parallel

    Word Parallel Bit Serial

    Department of CSE, ADCET, Ashta

  • 8/3/2019 Module 1-Part 1

    16/16

    Parallel Processing Applications

    Predictive Modeling and Simulations Numerical Weather Forecasting Oceanography and Astrophysics Socioeconomics and government use

    Engineering Design and Automation Finite Element Analysis Computational Aerodynamics Artificial Intelligence and Automation Remote Sensing Applications

    Medical Military and Basic Research Computer assisted tomography Genetic Engineering Weapon Research and Defense.

    Department of CSE, ADCET, Ashta