Scheduling Linearly Indexed Assignment Codes

It has been recently shown that linearly indexed Assignment Codes can be efficiently used for coding several problems especially in signal processing and matrix algebra. In fact, mathematical expressions for many algorithms are directly in the form of linearly indexed codes, and examples include the formulas for matrix multiplication, any m-dimensional convolution/correlation, matrix transposition, and solving matrix Lyapunov's equation. Systematic procedures for converting linearly indexed Assignment Codes to localized algorithms that are closely related to Regular Iterative Algorithms (RIAs) have also been developed. These localized algorithms can be often efficiently scheduled by modeling them as RIAs; however, it is not always efficient to do so. In this paper we shall analyze and develop systematic procedures for determining efficient schedules directly for the linearly indexed ACs and the localized algorithms. We shall also illustrate our procedures by determining schedules for examples such as matrix transposition and Gauss-Jordan elimination algorithm.

[1]  S. Kung,et al.  VLSI Array processors , 1985, IEEE ASSP Magazine.

[2]  Thomas Kailath,et al.  Subspace scheduling and parallel implementation of non-systolic regular iterative algorithms , 1989, J. VLSI Signal Process..

[3]  Thomas Kailath,et al.  Regular iterative algorithms and their implementation on processor arrays , 1988, Proc. IEEE.

[4]  V. van Dongen,et al.  Uniformization of linear recurrence equations: a step toward the automatic synthesis of systolic arrays , 1988, [1988] Proceedings. International Conference on Systolic Arrays.

[5]  Peter R. Cappello,et al.  Scheduling a system of affine recurrence equations onto a systolic array , 1988, [1988] Proceedings. International Conference on Systolic Arrays.

[6]  Richard M. Karp,et al.  The Organization of Computations for Uniform Recurrence Equations , 1967, JACM.

[7]  Thomas Kailath,et al.  Derivation, extensions and parallel implementation of regular iterative algorithms , 1989 .

[8]  Ed F. Deprettere,et al.  Converting sequential iterative algorithms to recurrent equations for automatic design of systolic arrays , 1988, ICASSP-88., International Conference on Acoustics, Speech, and Signal Processing.

[9]  Y. Wong,et al.  Broadcast removal in systolic algorithms , 1988, [1988] Proceedings. International Conference on Systolic Arrays.

[10]  D.I. Moldovan,et al.  On the design of algorithms for VLSI systolic arrays , 1983, Proceedings of the IEEE.

[11]  Dan I. Moldovan,et al.  Data broadcasting in linearly scheduled array processors , 1984, ISCA '84.