LAPACK Performance Libraries

Opti­mized LAPACK and BLAS libraries for Infi­neon AURIXTM
micro­con­trollers

The LAPACK Per­for­mance Libraries offer opti­mized imple­men­ta­tions of LAPACK (Lin­ear Alge­bra Pack­age) and BLAS (Basic Lin­ear Alge­bra Sub­pro­grams) for Infi­neon AURIXTM micro­con­trollers. These libraries sup­port sin­gle-pre­ci­sion oper­a­tions and are com­pat­i­ble with MATLAB and Simulink.

The LAPACK Libraries can be ben­e­fi­cial in ADAS-Cam­era and Motor Con­trol algo­rithms which require com­plex math­e­mat­ics to solve multi-vari­able equa­tions or move large amounts of data.

Prod­uct Brief
Free TRIAL
Request Quote


Performance

High­ly opti­mized for Infi­neon AURIXTM micro­con­trollers.
Ensures bet­ter per­for­mance for math­e­mat­i­cal com­pu­ta­tions.
Enhances the effi­cien­cy of Embed­ded sys­tems.


Reliability

Ensures reli­able and accu­rate math­e­mat­i­cal oper­a­tions.
Sup­port com­pli­ance with ISO 26262 up to ASIL B.
Opti­mized for develop­ment flex­i­bil­i­ty and effi­cien­cy.


Cost-Effective
Development

Enhance the effi­cien­cy and reli­a­bil­i­ty of math­e­mat­i­cal com­pu­ta­tions.
Designed for com­pat­i­bil­i­ty with the TASKING VX-toolset for Tri­Core.
Eas­i­ly portable across AURIXTM plat­forms.

High Accuracy

Sin­gle pre­ci­sion sup­port for BLAS & LAPACK func­tions includ­ed for even faster per­for­mance.
High­ly opti­mized numer­i­cal library.
Bit-exact Tri­Core result sim­u­la­tion on PCs avail­able from TASKING.

Flexible and Easy to Use

Direct­ly sup­port­ed by C/C++, MATLAB/Simulink, BASELABS and other model based pack­ages.
Improved LAPACK per­for­mance trans­lates direct­ly into appli­ca­tion per­for­mance improve­ments.
Eas­i­ly port exist­ing math­e­mat­i­cal soft­ware to embed­ded devices.
C source code avail­able upon request.
Easy deploy­ment in exist­ing soft­ware develop­ment envi­ron­ments via com­mand-line batch mode.

Applications that typically rely on, and benefit from LAPACK include:

ADAS Applications, e.g. some deterministic implementations of lane detection [4]

Solv­ing sys­tems of lin­ear equa­tions effi­cient­ly (Cholesky Decom­po­si­tion, …).
Coor­di­nate trans­for­ma­tions (Cam­era space to Carte­sian space).
Splines (dis­play rec­og­nized lanes).
Radar Appli­ca­tions.

MATLAB/Simulink based applications with code that calls into LAPACK/BLAS for

Matrix Mul­ti­pli­ca­tion.
Vec­tor Matrix Prod­ucts.

TASKING Prod­uct Doc­u­men­ta­tion con­sists of prod­uct release notes, user doc­u­men­ta­tion, prod­uct patch­es (where avail­able), webi­na­rs and prod­uct relat­ed infor­ma­tion.

https://www.tasking.com/support/lapack-performance-libraries/

FAQ

→ See all FAQ

Scroll to Top