SIMD processor arrays for image and video processing

A review

Thinh M. Le, W. M. Snelgrove, Sethuraman Panchanathan

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

SIMD processor arrays are becoming popular for their fast parallel executions of low- to medium-complexity image and video processing algorithms, and most stages of the compression standards (JPEG, H.263, and MPEG's). In many existing techniques, visual data processing algorithms and compression standards possess a high degree of parallelism. In particular, the processing of a certain pixel/block does not generally require data from a distant pixel/block, and the instructions for processing these pixels/blocks are usually identical. Thus, these algorithms map naturally onto the architecture of the SIMD processor arrays. In this paper, the architectures of the recently SIMD processor arrays will be reviewed together with algorithms demonstrating their superior features. Due to the length of the paper, only processor arrays implemented at the chip-level are discussed, especially those whose logic circuits are merged/embedded in the SRAM or DRAM memory process. While some processor arrays are designed by embedding the memory modules onto the existing processors, a significant number of processor arrays are realized by integrating logic circuits onto existing RAM (logic-in-memory) to save the inherently large memory bandwidth, and thus achieving a performance in the order of Tera instructions per second.

Original languageEnglish (US)
Pages (from-to)30-41
Number of pages12
JournalUnknown Journal
Volume3311
DOIs
StatePublished - 1998
Externally publishedYes

Fingerprint

SIMD (computers)
Parallel processing systems
image processing
central processing units
Processing
Data storage equipment
Pixels
Logic circuits
logic circuits
pixels
Integrating circuits
education
Dynamic random access storage
Static random access storage
Random access storage
embedding
logic
Bandwidth
modules
chips

Keywords

  • Logic-in-memory
  • Parallel processing
  • Processor arrays
  • SIMD architecture

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Condensed Matter Physics

Cite this

SIMD processor arrays for image and video processing : A review. / Le, Thinh M.; Snelgrove, W. M.; Panchanathan, Sethuraman.

In: Unknown Journal, Vol. 3311, 1998, p. 30-41.

Research output: Contribution to journalArticle

Le, Thinh M. ; Snelgrove, W. M. ; Panchanathan, Sethuraman. / SIMD processor arrays for image and video processing : A review. In: Unknown Journal. 1998 ; Vol. 3311. pp. 30-41.
@article{7503024e3d3c4ed9b364ee6c26c91147,
title = "SIMD processor arrays for image and video processing: A review",
abstract = "SIMD processor arrays are becoming popular for their fast parallel executions of low- to medium-complexity image and video processing algorithms, and most stages of the compression standards (JPEG, H.263, and MPEG's). In many existing techniques, visual data processing algorithms and compression standards possess a high degree of parallelism. In particular, the processing of a certain pixel/block does not generally require data from a distant pixel/block, and the instructions for processing these pixels/blocks are usually identical. Thus, these algorithms map naturally onto the architecture of the SIMD processor arrays. In this paper, the architectures of the recently SIMD processor arrays will be reviewed together with algorithms demonstrating their superior features. Due to the length of the paper, only processor arrays implemented at the chip-level are discussed, especially those whose logic circuits are merged/embedded in the SRAM or DRAM memory process. While some processor arrays are designed by embedding the memory modules onto the existing processors, a significant number of processor arrays are realized by integrating logic circuits onto existing RAM (logic-in-memory) to save the inherently large memory bandwidth, and thus achieving a performance in the order of Tera instructions per second.",
keywords = "Logic-in-memory, Parallel processing, Processor arrays, SIMD architecture",
author = "Le, {Thinh M.} and Snelgrove, {W. M.} and Sethuraman Panchanathan",
year = "1998",
doi = "10.1117/12.304673",
language = "English (US)",
volume = "3311",
pages = "30--41",
journal = "Scanning Electron Microscopy",
issn = "0586-5581",
publisher = "Scanning Microscopy International",

}

TY - JOUR

T1 - SIMD processor arrays for image and video processing

T2 - A review

AU - Le, Thinh M.

AU - Snelgrove, W. M.

AU - Panchanathan, Sethuraman

PY - 1998

Y1 - 1998

N2 - SIMD processor arrays are becoming popular for their fast parallel executions of low- to medium-complexity image and video processing algorithms, and most stages of the compression standards (JPEG, H.263, and MPEG's). In many existing techniques, visual data processing algorithms and compression standards possess a high degree of parallelism. In particular, the processing of a certain pixel/block does not generally require data from a distant pixel/block, and the instructions for processing these pixels/blocks are usually identical. Thus, these algorithms map naturally onto the architecture of the SIMD processor arrays. In this paper, the architectures of the recently SIMD processor arrays will be reviewed together with algorithms demonstrating their superior features. Due to the length of the paper, only processor arrays implemented at the chip-level are discussed, especially those whose logic circuits are merged/embedded in the SRAM or DRAM memory process. While some processor arrays are designed by embedding the memory modules onto the existing processors, a significant number of processor arrays are realized by integrating logic circuits onto existing RAM (logic-in-memory) to save the inherently large memory bandwidth, and thus achieving a performance in the order of Tera instructions per second.

AB - SIMD processor arrays are becoming popular for their fast parallel executions of low- to medium-complexity image and video processing algorithms, and most stages of the compression standards (JPEG, H.263, and MPEG's). In many existing techniques, visual data processing algorithms and compression standards possess a high degree of parallelism. In particular, the processing of a certain pixel/block does not generally require data from a distant pixel/block, and the instructions for processing these pixels/blocks are usually identical. Thus, these algorithms map naturally onto the architecture of the SIMD processor arrays. In this paper, the architectures of the recently SIMD processor arrays will be reviewed together with algorithms demonstrating their superior features. Due to the length of the paper, only processor arrays implemented at the chip-level are discussed, especially those whose logic circuits are merged/embedded in the SRAM or DRAM memory process. While some processor arrays are designed by embedding the memory modules onto the existing processors, a significant number of processor arrays are realized by integrating logic circuits onto existing RAM (logic-in-memory) to save the inherently large memory bandwidth, and thus achieving a performance in the order of Tera instructions per second.

KW - Logic-in-memory

KW - Parallel processing

KW - Processor arrays

KW - SIMD architecture

UR - http://www.scopus.com/inward/record.url?scp=0032374261&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032374261&partnerID=8YFLogxK

U2 - 10.1117/12.304673

DO - 10.1117/12.304673

M3 - Article

VL - 3311

SP - 30

EP - 41

JO - Scanning Electron Microscopy

JF - Scanning Electron Microscopy

SN - 0586-5581

ER -