Synthesis of pipelined DSP accelerators with dynamic scheduling
- 19 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
Abstract
To construct complete systems on silicon, application specific DSP accelerators are needed to speed up the execution of high throughput DSP algorithms. In this paper, a methodology is presented to synthesize high throughput DSP functions into accelerator processors containing a datapath of highly pipelined, bit-parallel hardware units. Emphasis is put on the definition of a controller architecture that allows efficient run-time schedules of these DSP algorithms on such highly pipelined data paths. The methodology is illustrated by means of an FFT butterfly accelerator block.Keywords
This publication has 9 references indexed in Scilit:
- Scheduling and hardware sharing in pipelined data pathsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- Design of a 30 MHz, 32/16/8-points DCT processor with PhideoPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Design of a pipelined datapath synthesis system for digital signal processingIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 1994
- Cathedral-IIIPublished by Association for Computing Machinery (ACM) ,1991
- Sehwa: a software package for synthesis of pipelines from behavioral specificationsIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 1988
- Synchronous data flowProceedings of the IEEE, 1987
- Present SLAC Accelerator Computer Control System FeaturesIEEE Transactions on Nuclear Science, 1981
- Pipeline ArchitectureACM Computing Surveys, 1977
- An algorithm for the machine calculation of complex Fourier seriesMathematics of Computation, 1965