HPC - Universidad de Sevilla


Linear Algebra


The BLAS (Basic Linear Algebra Subprograms) are routines that provide standard building blocks for performing basic vector and matrix operations

See Netlib for the "original".


CBLAS is not a standalone library, but an interface to the FORTRAN BLAS routines. It can be found at Netlib, most BLAS implementation come with a CBLAS interface.


LAPACK provides routines for solving systems of simultaneous linear equations, least-squares solutions of linear systems of equations, eigenvalue problems, and singular value problems. LAPACK uses BLAS routines extensively, so LAPACK preformance is closely related to BLAS performance.

See Netlib for the "original".


Highly optimized implementation of BLAS and LAPACK for different arquitectures, freely available. The best BLAS for the "poor" scientist. Includes LAPACK and CBLAS interface.

Intel MKL

Intel MKL (Math Kernel Library) is a highly optimized library for Intel processors. Includes BLAS, LAPACK, sparse and parallel versions, partial differential equations, FFT and vector mathematical functions. The best library out there for Intel, only available with Intel compiler products which start at 699$. Free versions for students, educators and researchers here.

More Info


AMD ACML (AMD Core Math Library) is AMD's answer to Intel's MKL. AMD has transitioned from propietary to open source, so these libraries are available for free. BLAS, LAPACK, FFT and RNG.