Document Number: 253425-009US
Revision 9
The Intel® Math Kernel Library (Intel® MKL) is a set of highly optimized
routines for solving problems of computational linear algebra,
performing the discrete Fourier transforms, and solving some other computation intensive problems.
The library includes routines of the BLAS, Sparse BLAS, LAPACK and ScaLAPACK1) packages (Fortran interfaces), Sparse Solver,
Interval Linear Solvers, CBLAS (C interface to BLAS routines),
as well as Discrete Fourier (with Cluster DFTI1)) and Fast Fourier transform2) routines, Vector Mathematical Functions, Vector Statistical Library (Fortran and C interfaces for random number
generators and convolution/correlation mathematical operations).The library functions ensure high performance when run on
Intel® processors. Level 3 BLAS and most LAPACK routines, in particular, take advantage of multiprocessor
computation.
Intel MKL implementation of GMP arithmetic functions includes arbitrary precision arithmetic operations on integer
numbers. The interfaces of such functions fully match the GNU Multiple Precision* (GMP) Arithmetic Library.
For specifications of these functions, please see
http://www.swox.com/gmp/manual/Integer-Functions.html.
Intel MKL along with Fortran-77 interface to BLAS, Sparse BLAS, and LAPACK routines also supports
Fortran-95 interface. This interface for BLAS, Sparse BLAS Level 1, and LAPACK routines uses simplified routine calls with
shorter argument lists. The calling sequence
for Fortran-95 interface is given in the syntax section of the routine description immediately after Fortran-77 calls.
The simplified calls are not implemented in the current versions of Sparse BLAS Level 2 and Level 3 routines.
This Quick Reference includes the following sections:
1)ScaLAPACK
software and Cluster DFTI of DFT functions are included only with the Intel® Cluster MKL product. To ensure a correct display of this document, use the following recommended browser versions:
Internet Explorer* 5.5 or higher (on Windows*), Netscape* 4.79 or Mozilla* 1.2.1 or higher (on Linux*).
For more information about the library, please refer to
Intel® Math Kernel Library Reference Manual (document number 630813-021US)
and the Web site at
www.intel.com/software/products/mkl/docs/manuals.htm.
Note.
If you do not see the Table of Contents in the left frame
while your browser seems to have "loaded everything", please
enable JavaScript.
2)The FFT functions have been deprecated
in Intel MKL and remain in the library only for legacy reasons. Use DFT functions instead.
* Legal Information © 1999, 2002-2006, Intel Corporation