Kane Computing Ltd - Home Page
Kane Computing Ltd
Javascript DHTML Drop Down Menu Powered by dhtml-menu-builder.com
Line Search:
Quick Product Finder
Products by Supplier
News
What's this?
Customer Success Stories
Press Releases
eNews (subscribe to enews)

Registered in England No: 4264387
Registered Office:
c/o Bulcock & Co. 10 The Bull Ring, Northwich, Cheshire, CW9 5BS, UK.

Kane Computing Ltd,
Suite I, Ascot Court,
71-73 Middlewich Road,
Northwich,
Cheshire, CW9 7BP
United Kingdom
Tel: +441606351006
Fax: +441606351007
Email: Sales-Support-Marketing
©Copyright: KCL 1986-2012

Federation of Small Businesses

GDD7000 - Hand Coded Optimised LINPACK Library

GDD Libraries

Features

  • Matrix decomposition (SVD, QR, Cholesky)
  • Linear systems solutions (Teoplitz, Hermitian, etc)
  • Memory allocation functions
  • Big Endian (BE) or Little Endian (LE) (as xx suffix to part number)
Ordering Info: Part No. GDD7000-xx

Description

GDD7000 is an extended C-callable LINPACK linear algebraic systems package for TMS320C6000.

The library is a set of about 100 functions and macros that solve algebraic systems of linear equations with various matrices, including; real or complex, general, band, symmetric, Toeplitz and others. Several types of matrix decompositions like SVD or QR can be performed by the library functions. The input matrices and right-hand side (RHS) vectors as well as the solution are set and found as IEEE-754 floating-point format numbers.

A set of memory allocation functions provides economical storage for general/structured matrix classes; minimizing the amount of RAM required to store structured matrices, such as symmetric, band and triangular classes of matrices.

The library is supported for use in any development environment using TI Code Generation Tools for the TMS320C6000 DSP and supports both big endian and little endian memory formats. All the functions are C callable and comply with TI's C environment calling conventions. The library implements and makes use of the complete Level 1 BLAS (Basic Linear Algebra Subroutines) standard.

The functions have been optimised algorithmically at assembly level to fully utilize advantages of TMS320C6000 parallel architecture, floating-point arithmetic and pipeline. BLAS Level 1 functions are hand-coded in assembly to obtain maximum possible performance from the TMS320C6000 DSPs.

Related Products

Line

Downloads

Datasheet (PDF 116KB)
Benchmarks (PDF 130KB)

 

Line