Title :
A cordic-based processor extension for scalar and vector processing
Author :
Farivar, R. ; Fazeli, M. ; Sarbazi-Azad, H.
Author_Institution :
Dept. of Comput. Eng., Sharif Univ. of Technol., Tehran, Iran
Abstract :
Trigonometric calculations play a significant role in scientific and multimedia applications. Such applications usually use vectors of data, thus it is essential to use SIMD-based computing for them. State of the art processors such as Pentium family processors have SIMD capabilities, MMX, SSE and SSE2 to name a few. But these extensions usually come short of performing trigonometric functions calculation. This paper presents a CORDIC-based vector processor extension to accelerate trigonometric calculations. The suggested architecture works with a vector of data, as well as scalar data. This architecture has been synthesized both for FPGA and ASIC and the results have shown a significant improvement in throughput over conventional processors.
Keywords :
application specific integrated circuits; digital arithmetic; field programmable gate arrays; parallel architectures; vector processor systems; ASIC; CORDIC-based vector processor extension; FPGA; SIMD-based computing; scalar processing; trigonometric calculation; vector processing; Application software; Application specific integrated circuits; Computer architecture; Coprocessors; Discrete Fourier transforms; Educational institutions; Field programmable gate arrays; Fourier transforms; Hardware; Throughput; ASIC; CORDIC Algorithm; FPGA; Pool of Processors; Processor Extension; SIMD; Trigonometric functions; Variable precision; Vector Processing;
Conference_Titel :
Parallel and Distributed Processing Symposium, 2005. Proceedings. 19th IEEE International
Print_ISBN :
0-7695-2312-9
DOI :
10.1109/IPDPS.2005.11
