A Design Methodology for Very Large Array Processors—part 1: Gipop Processor Array

Venkateswaran, N.; Pattabiraman, Srividhya; Devanathan, R.; Kumaran, Bharath; Ahmed, Abdulrahman A.; NARAYANAN, SANKARA; RADHARAMANAN,

doi:10.1142/9789812797766_0004

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The basic technique used for designing a massively parallel architecture for Very Large Problems (VLPs) is to express the VLP's in terms of G-set functions [5] corresponding to Sum of Product (SOP), Inner Product (IP), Product of Sum (POS),…”

Section: Pacube Vlsi Array Based Massively Parallel Architecture For mentioning

confidence: 99%

“…These arrays are ultrahigh performance chips meant for special purpose systems. Another important feature is the architecture of the macrocell, which is designed in such a way that the functional units corresponding to the G-set equations [5] when mapped on the macrocell arrays possess identical data flow control. This leads to highly simplified control design for executing complex computations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Recursive Gabor transform for images of very large size and design of massively parallel architecture

Venkateswaran

Venkatachalam²,

Swaminathan³

et al.

Proceedings Fourth International Conference on High-Performance Computing

Self Cite

View full text Add to dashboard Cite

The conventional Gabor Transforms, Complete and Generalized cannot be appliedfor image matrices received as different partitions. In practice, very large images are received in the form of submatrix partitions. Then, taking transform on different partitions (of the same image) at different times is of prime importance. The situation can be easily handled i f the transformation algorithm is made recursive. In this paper we present a Recursive Gabor Transform (RGT) amenable for pipelining panitions of diflerent images interleaved. Apriori knowledge of image size is not required for the RGT unlike the conventional Gabor Transform. A dedicated rnassively parallel architecture based on Pacube VLSI arrays is presented which achieves a high degree of pipelining with respect to different image matrices. The pipelining rate is proportional to the delay of a Pacube macrocell in mode I operation.where T, T > 0. When 'T = T, the critical sampling case (1) is the Comp1el.e Gabor Transform (CGT). When T c T', the oversampling case, (1) is the GGT, g is the window function. The matrix formulation proposed by Yao [ 11, for GGT can be expressed as f = G a ... (2) where f = (fo, f,, ..., fKM.I)T and a = (a, , , a,, ..,, +M.I)T We can then decompose the coefficient vector a = (IK * E,) D 1 f ... (3) where I, is the Kronecker product, the form is a matrix of euv = exp (-2mv/M'), U, v = 0,1, ,.. , M'-1

show abstract

Section: Pacube Vlsi Array Based Massively Parallel Architecture For mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Recursive Gabor transform for images of very large size and design of massively parallel architecture

Venkateswaran

Venkatachalam²,

Swaminathan³

et al.

Proceedings Fourth International Conference on High-Performance Computing

Self Cite

View full text Add to dashboard Cite

show abstract

A Design Methodology for Very Large Array Processors—part 1: Gipop Processor Array

Cited by 1 publication

References 0 publications

Recursive Gabor transform for images of very large size and design of massively parallel architecture

Recursive Gabor transform for images of very large size and design of massively parallel architecture

Contact Info

Product

Resources

About