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

Name of all sap databases, All SAP Databases are Relational Databases ...

All SAP Databases are Relational Databases is the all sap databases.

Difference between a null pointer and a void pointer, A NULL pointer is a p...

A NULL pointer is a pointer of any type whose value is zero. A void pointer is a pointer to an object of an unknown type, and is guaranteed to have sufficient bits to hold a pointe

Diffeomorphism, A different smooth structure on R: Show that (U, f) given b...

A different smooth structure on R: Show that (U, f) given by U = R, f : x -> x3, is a local chart of the topological manifold M = R which is not a member of the standard smoo

What is functions of input - output interface, Q. What is Functions of Inpu...

Q. What is Functions of Input - Output Interface? An I/O interface is bridge between processor and I/O devices. It controls data exchange between external devices and main mem

C, solution for oadovan string inc language

solution for oadovan string inc language

Explain parallelism based on grain size in detail, Parallelism based on Gra...

Parallelism based on Grain size Grain size : Grain size/ Granularity are a measure that defines how much computation is involved in a process. Grain size is concluded by count

Explain 100 line exchange with two-motion line finders, Explain 100 line ex...

Explain 100 line exchange with two-motion line finders. Design: In such design, here are 24 line finders. When any of the 100 subscribers has to find access to any of the 24

How deep does fifo require to be stop underflow or overflow, Given the subs...

Given the subsequent FIFO and rules, how deep does the FIFO require to be to stop underflow or overflow? RULES: a. frequency(clk_A) = frequency(clk_B) / 4 b. per

State the concept of hierarchy in object oriented, Hierarchy Hierarchy...

Hierarchy Hierarchy is a process of ranking or ordering of abstraction. Hierarchy is decided by the help of 'divide and conquer' principle. Complex objects can be described

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