Briefly explain array processing in detail, Computer Engineering

Assignment Help:

Array Processing

We have seen that for executing vector operations the pipelining conception has been used. There is other method for vector operations. If we have array of n processing elements (PEs) It implies that several ALUs for storing several operands of vector then an n instruction i.e. vector addition is broadcast to all PEs , such that they sum all operands of vector at the same instance which means all PEs will perform computation in parallel. All PEs are synchronised under one control unit. This organisation of synchronous array of PEs for vector operations is termed as Array Processor. An array processor can handle one instruction and numerous data streams as we have seen in the case of SIMD organisation. So array processors are called SIMD array computers too. 

The organisation of an array processor is displayed in Figure below. The subsequent elements are organised in an array processor:

672_Array Processing.png

Figure: Organisation of SIMD Array Processor

Control Unit (CU): All PEs are under control of singl0065 control unit. CU manages the inter communication between the PEs. There is a local memory of control unit which is known as CY memory. The user programs are loaded in the CU memory. The vector instructions in program are decoded by CU and transmit to array of PEs. Instruction decoding and fetch is done by CU only. 

 

Processing elements (PEs): Every processing element comprises of ALU, its registers and a local memory for storage of distributed data. This PEs has been interconnected through an interconnection network. All PEs receive instructions from the CU and several element operands are fetched from their local memory. So all PEs execute the similar function synchronously in a lock-step fashion under the control of CU.   

 

It might be probable that all PEs require not participating in the execution of a vector instruction. So it is needed to adopt a masking technique to control the status of every PE. A masking vector is used to manage the status of all PEs such that only enabled PEs are permitted to take part in execution and others are disabled.

 

Interconnection Network (IN): IN executes data exchange among PEs and data routing and manipulation functions. This IN is under control of Control Unit.

 

Host Computer: An array processor can be connected to a host computer via the CU. The objective of the host computer is to broadcast a series of vector instructions from CU to the PEs. So host computer is a general-purpose machine which acts as a manager of entire system.

Array processors are "Special purpose computers" that have been used for the following:

  • real-time scene analysis
  • matrix algebra
  • matrix eigen value calculations
  • various scientific applications

 

SIMD array processor on large scale has been developed by NASA for earth resources satellite image processing. This machine has been named "Massively parallel processor" (MPP) since it comprises 16,384 processors which work concurrently. MPP provides real-time time varying scene analysis. However array processors aren't commercially popular as well as aren't commonly used. The reasons are that array processors are complicated to program in comparison to pipelining and there is a problem in vectorization.


Related Discussions:- Briefly explain array processing in detail

Feasebility study, what are the feasibility study of online result manageme...

what are the feasibility study of online result management system?

Example of time division multiplexing, Two computers using TDM take up turn...

Two computers using TDM take up turns to send 100-bytes packet over a shared channel that operates at 64000 bits per second. The hardware takes 100 microseconds after one computer

What are the different methods used for handling, What are the different me...

What are the different methods used for handling the situation when multiple interrupts occurs? 1) Vectores interrupts 2) Interrupt nesting 3) Simultaneous Requests.

Programming with parallel virtual machine, Q. Programming with parallel vir...

Q. Programming with parallel virtual machine? The general method for writing a program with PVM is like this:  A user writes one or more sequential programs in C++, C or FOR

Explain the transport layer in detail, Explain the transport layer in detai...

Explain the transport layer in detail. Transport Layer: The transport layer controls and makes sure the end-to-end integrity of the data message propagated by the network am

Explain about working of multiplexer, Q. Explain about working of Multiplex...

Q. Explain about working of Multiplexer? Multiplexer is one of the fundamental building units of a computer system that in principle permits sharing of a common line by more th

What is the most common biometrics, What is the most common biometrics? ...

What is the most common biometrics? The most common biometrics is the given below: Face geometry (Photo): In this the computer takes the picture of your face and matches t

C program, program for finding the area under the curve   #includ...

program for finding the area under the curve   #include float start_point, /* GLOBAL VARIABLES */ end_point, total_area; int

Explain procedure level of parallel processing, Procedure Level Here, p...

Procedure Level Here, parallelism is obtainable in the form of parallel executable procedures. In that case, design of algorithm plays a key role. E.g. every thread in Java is

Write Your Message!

Captcha
Free Assignment Quote

Assured A++ Grade

Get guaranteed satisfaction & time on delivery in every assignment order you paid with us! We ensure premium quality solution document along with free turntin report!

All rights reserved! Copyrights ©2019-2020 ExpertsMind IT Educational Pvt Ltd