 |
» |
|
|
|
CXML is a collection of mathematical routines optimized for Alpha systems. These subroutines perform numerically intensive operations that occur frequently in engineering and scientific computing, such as linear algebra and signal processing. CXML can help reduce the cost of computation, enhance portability, and improve productivity.
CXML is available for free download from the Math Libraries web site. It is also included in HP Fortran, Visual Fortran, and HP Fortran for Linux Alpha.
For more information on CXML, click on any of the following topics:
|
|
|
|
|
For the very latest information on CXML and CPML (our Portable Math Library for Alpha), or to download either library for free, visit the Math Libraries web site.
|
|
|
|
|
|
- Productivity - CXML provides over 1000 callable functions, allowing users to spend more time on the actual problem and less time worrying about numerical analysis.
- Performance - In CXML, key routines have been extensively optimized for the Alpha architecture. In addition, many routines have been parallelized to take advantage of Alpha SMP servers.
- Reliability - CXML routines are based on the best algorithms and numerical techniques. Each new release is run through an extensive test suite to verify correctness.
|
|
|
|
|
|
|
CXML is a set of scientific subroutines optimized for Alpha systems, organized into four libraries:
| BLAS - Basic Linear Algebra Subroutines |
| LAPACK - An extensive library of Linear System and Eigenproblem Solvers |
| Sparse Linear System Solvers - A library of direct and iterative sparse solvers |
| Signal Processing - FFTs, sine/cosine transforms, convolution, correlation and filters |
|
|
|
|
|
|
The Basic Linear Algebra Library includes the industry-standard Basic Linear Algebra Subprograms (BLAS) Level 1, Level 2, and Level 3. Also included are subprograms for BLAS Level 1 Extensions, Sparse BLAS Level 1, and Array Math Functions (VLIB).
|
|
The LAPACK library |
|
|
The Linear System and Eigenproblem Solver Library provides the complete LAPACK package, compiled, tested, and ready to use. LAPACK is a new, industry-standard package offering an extensive set of linear system and eigenproblem solvers. LAPACK uses blocked algorithms that are better suited to Alpha's advanced 64-bit architecture. LAPACK will supersede LINPACK and EISPACK for most users. For more information on LAPACK, click here.
|
|
The sparse solver library |
|
|
The Sparse Linear System Library provides both direct and iterative sparse linear system solvers. The direct solver package supports both symmetric and nonsymmetric sparse matrices stored using the skyline storage scheme. The iterative solver package contains a basic set of storage schemes, preconditioners, and iterative solvers. The design of this package is modular and matrix-free, allowing future expansion and easy modification by users.
|
|
The signal processing library |
|
|
The Signal Processing Library provides a basic set of signal processing functions. Included are one-, two-, and three-dimensional Fast Fourier Transforms (FFT), group FFTs, cosine/sine transforms, convolution, correlation, and digital filters.
|
|
|
|
|
|
An advantage of CXML's support for industry standard interfaces is that other products can use CXML to improve performance on Alpha systems. For example, the following math libraries can use the optimized CXML BLAS routines, resulting in better performance:
- Commercial software that can call optimized CXML BLAS:
- Fast Matrix Solver (FMSLIB) from Multipath Corporation is a package of routines to solve very large, out-of-core sparse problems. For more information, visit the FMSLIB web site.
- The Alpha versions of MATLAB from Mathworks can now take advantage of the CXML BLAS.
- IMSL Math Library from Visual Numerics
- NAG Libraries from Numerical Algorithms Group.
- Sources of public domain and commercial numerical software:
- Netlib maintains a large library of numerical software and is the source for the public domain versions of BLAS and LAPACK.
- GAMS (Guide to Available Mathematical Software) is a repository of information on mathematical software, both public domain and commercial.
|
|
|
|
|
Printable version
