Explain parallelism based on grain size in detail, Computer Engineering

Assignment Help:

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 counting number of instructions in a program segment. The subsequent types of grain sizes have been recognized (shown in Figure):

406_Parallelism based on Grain size.png

Figure: Types of Grain sizes

1)  Fine Grain: This type includes nearly less than 20 instructions.

2)  Medium Grain: This type includes nearly less than 500 instructions.

3)  Coarse Grain: This type includes nearly greater than or equal to one thousand instructions.

Based on these grain sizes, parallelism may be classified at several stages in a program. These parallelism stages create a hierarchy according to which, lower the level the finer is granularity of process. The amount of parallelism reduces with raise in level. Each level according to a grain size requires scheduling overhead and communication. Following are parallelism levels (shown in Figure):

408_Parallelism based on Grain size 1.png

Figure: Parallelism Levels

1)  Instruction level: It is the lowest level and degree of parallelism is highest at this level. Fine grain size is used at statement or instruction level as just few instructions make the grain size here. The fine grain size may perhaps vary according to type of the program. E.g. for scientific applications, Instruction level grain size may be higher. As the higher degree of parallelism is able to be achieved at this level, the overhead for a programmer would be more.

2)  Loop Level: This is other stage of parallelism where iterative loop instructions able to be parallelized. Fine grain size is used at this stage too. Simple loops in program are simple to parallelize whereas the recursive loops are hard. This kind of parallelism can be achieved by the compilers.

 3) Subprogram or Procedure Level: This level consists of subroutines, subprograms or procedures. Medium grain size is used at this level including several thousands of instructions in a process. Multiprogramming is applied at this stage. Parallelism at this level has been developed by programmers however not through compilers. Parallelism through compilers hasn't been attained at the medium and coarse grain size.

4)  Program Level: It is the last level consisting of independent programs for parallelism. Coarse grain size is used at this stage including tens of thousands of instructions. Time sharing is attained at this level of parallelism. Parallelism at this stage has been exploited through the operating system.  

The relation between parallelism levels and grain sizes has been shown in Table.

628_Parallelism Levels.png

Table: Relation between grain sizes and parallelism

Coarse grain parallelism is conventionally applied in shared memory or tightly coupled multiprocessors such as the Cray Y-MP. Loosely coupled systems are used to perform medium grain program segments. Fine grain parallelism has been monitored in SIMD organization of computers.


Related Discussions:- Explain parallelism based on grain size in detail

Decoding of slightly encoded micro-instructions, Q. Decoding of slightly en...

Q. Decoding of slightly encoded micro-instructions? Generally micro-programmed control unit designs are neither totally unencoded nor highly encoded. They are slightly coded. T

#project, how to build PVR set-top box using raspberry

how to build PVR set-top box using raspberry

Name the languages introduced in second generation, Name the languages intr...

Name the languages introduced in second  generation During  the time of  second  generation  many  high  level  programming  languages  were introduced, such as FORTRAN (1956),

Analysis of algorithms, Analysis of Algorithms For this task, each stu...

Analysis of Algorithms For this task, each student should do two things: An empirical analysis of the runtime and comparisons made for all algorithms as a function of input si

Architecture of artificial neural networks, Architecture of Artificial neur...

Architecture of Artificial neural networks: Presumably "Artificial Neural Networks" consist of a number of units that are mini calculation devices. But they take in real-value

Java, differentiate between multitasking and multithreading in java

differentiate between multitasking and multithreading in java

What is the use of fork and exec system calls, What is the use of fork and ...

What is the use of fork and exec system calls?  Fork is a system call by which a new process is formed. Exec is also a system call, which is used after a fork by one of the two

Btree, What should the size of ''t'' in btree be depending on the hard disk...

What should the size of ''t'' in btree be depending on the hard disk size

Graph has a minimum spanning tree already computed, Assume that a graph has...

Assume that a graph has a minimum spanning tree already computed.  How fastly can the minimum spanning tree be updated if a new vertex and incident edges are added to G? If the

What is public, * Public, protected and private are 3 access specifier in C...

* Public, protected and private are 3 access specifier in C++. * Public data members and member functions are accessible outside the class. * Protected data members and memb

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