liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Implementation of Elementary Functions for a Fixed Point SIMD DSP Coprocessor
Linköping University, Department of Electrical Engineering, Computer Engineering.
2010 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

This thesis is about implementing the functions for reciprocal, square root, inverse square root and logarithms on a DSP platform.

A multi-core DSP platform that consists of one master processor core and several SIMD coprocessor cores is currently being designed by a team at the Computer Engineering Department of Linköping University.

The SIMD coprocessors’ arithmetic logic unit (ALU) has 16 multipliers to support vector multiplication instructions. By efficiently using the 16 multipliers, it is possible to evaluate polynomials very fast. The ALU does not have (hardware) support for floating point arithmetic, so the challenge is to get good precision by using fixed point arithmetic.

Precise and fast solutions to implement the mathematical functions are found by converting the fixed point input to a soft floating point format before polynomial approximation, choosing a polynomial based on an error analysis of the polynomial approximation, and using Newton-Raphson or Goldschmidt iterations to improve the precision of the polynomial approximations.

Finally, suggestions are made of changes and additions to the instruction set architecture, in order to make the implementations faster, by efficiently using the currently existing hardware.

Place, publisher, year, edition, pages
2010. , 134 p.
Keyword [en]
SIMD, DSP, mathematical functions, elementary functions, polynomial approximation, fixed-point arithmetic
National Category
Computer Engineering
URN: urn:nbn:se:liu:diva-63576ISRN: LiTH-ISY-EX--10/4399--SEOAI: diva2:380664
2010-12-10, 10:15 (English)
Available from: 2011-01-03 Created: 2010-12-22 Last updated: 2011-02-02Bibliographically approved

Open Access in DiVA

fulltext(701 kB)634 downloads
File information
File name FULLTEXT01.pdfFile size 701 kBChecksum SHA-512
Type fulltextMimetype application/pdf

By organisation
Computer Engineering
Computer Engineering

Search outside of DiVA

GoogleGoogle Scholar
Total: 634 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 335 hits
ReferencesLink to record
Permanent link

Direct link