Feng’s classification

Tse-yun Feng suggests the use of degree of parallelism to classify various computer architecture . It is based on sequential and parallel operations at a bit and word level. [1]

About degree of parallelism

Maximum degree of parallelism

The maximum number of binary digits is called the maximum parallelism degree. If a processor is processing P bits in unit time, then P is called the maximum degree of parallelism . [2]

Average degree of parallelism

Let i = 1, 2, 3, …, T be the different timing instants and P 1 , P 2 , …, P T be the corresponding bits processed. Then,

 Average parallelism, P a = (P 1 + P 2 + ... P T ) / T

Processor utilization

Processor Utilization, μ = P a / P

The maximum degree of parallelism depends on the structure of the Arithmetic and Logic Unit. Higher degree of parallelism indicates a highly parallel ALU or processing element. Average parallelism depends on the hardware and the software. Higher average parallelism can be achieved through concurrent programs.

Types of classification

According to Feng’s classification, computer architecture can be classified into four. The classification is based on the content stored in memory are processed. The contents can be the data or instructions. [3]

  1. Word Serial Bit Serial (WSBS)
  2. Word Serial Parallel Bit (WSBP)
  3. Word Parallel Bit Serial (WPBS)
  4. Word Parallel Parallel Bit (WPBP)

Word Serial Bit Serial (WSBS)

One bit of one word is processed at a time. This represents the maximum processing time and processing time.

Word Serial Bit Parallel (WSBP)

found. found…………………. All bits of a word are processed at a time. Parallel bit all bits of a word.

Word Parallel Bit Serial (WPBS)

It has been called a bit slice processing because it is processes at a time. Word parallel signifies selection of all words. It can be considered as a bit of all words are processed at a time.

Word parallel bit parallel (WPBP)

It is known as fully parallel processing in which an array is nxm bits is processes at one time. Maximum parallelism is achieved here.

Limitations of Feng’s classification

It fails to project the concurrency in pipeline processors, as it does not account for concurrency handle by pipe-lined design.

See also

  • Flynn’s taxonomy


  1. Jump up^ http://www.ijcaonline.org/archives/volume148/number14/thakur-2016-ijca-911285.pdf
  2. Jump up^ Salih, Mariam A. Salih. “Architectural Classification” (PDF) . Retrieved 5 October 2016 .
  3. Jump up^ https://www.gehucsit.files.wordpress.com/2013/09/handlers-classsification1.pdf

Leave a Reply

Your email address will not be published. Required fields are marked *

Copyright computerforum.eu 2018
Shale theme by Siteturner