Mathematical & Other Libraries

Bright Cluster Manager® includes a comprehensive list of mathematical and other programming libraries.

Mathematical Libraries

Examples of mathematical libraries available with Bright Cluster Manager are provided below. Some libraries may require purchase of a license.

• ACML — The AMD Core Math Library (ACML) contains the complete set of LAPACK routines, the complete set of functions from the Basic Linear Algebra Subprograms (BLAS) and the extended BLAS (sparse). It also includes a small set of BLAS-like functions to compute minima for various data types and Fast Fourier Transform (FFTA Fast Fourier transform (FFT) is an efficient algorithm to compute the Discrete Fourier Transform (DFT) and its inverse.) routines. The ACML is optimized for and supports all AMD processors, but also has good performance on Intel processors.
• FFTW — The Fastest Fourier Transform in the West (FFTW) library is a highly optimized routine for fast fourier transforms with runtime algorithm planning and code generation. It supports fast fourier transforms and reverse fast fourier transforms for single, double and extended precision floating point numbers, both for real and complex numbers. FFTW is open source.
• GMP — The GNU MultiPrecision Library (GMP) is a library for arbitrary precision arithmetic, signed integer operations, rational numbers and floating point numbers. GMP is designed for speed, for both small and very large operands. GMP is fast because it uses fullwords as the basic arithmetic type, it uses fast algorithms, it carefully optimizes assembly code for many CPUs' most common inner loops, and it generally emphasizes speed over simplicity/elegance in its operations. In Bright Cluster Manager, GMP is patched with assembler-tuned routines for extra performance. GMP is open source.
• GotoBLAS & ATLAS — The GotoBLAS and ATLAS (Automatically Tuned Linear Algebra Software) libraries are highly optimized implementations of the BLAS (Basic Linear Algebra Subprograms) linear algebra kernel library. License restrictions may apply to the GOTO libraries.
• MKL — The Intel Math Kernel Library (MKL) contains the complete set of LAPACK routines, the complete set of functions from the basic linear algebra subprograms (BLAS), the extended BLAS (sparse). It also includes a small set of BLAS-like functions to compute minima for various data types and Fast Fourier Transform (FFT) routines. The MKL is optimized for and supports all Intel processors, including the Itanium2. MKL requires purchase of a license.
• ScaLAPACK — Scalable LAPACK (ScaLAPACK), is a library of high performance linear algebra routines specifically developed for distributed memory message-passing MIMDMIMD (Multiple Instruction stream, Multiple Data stream) is a technique employed to achieve parallelism. Machines using MIMD have a number of processors that function asynchronously and independently. computers which use MPIMessage Passing Interface and/or PVMParallel Virtual Machine, such as clusters. ScaLAPACK in Bright Cluster Manager is optimized to work with fast interconnects like Gigabit, Myrinet and InfiniBand.

Other Libraries

Examples of other programming libraries included with Bright Cluster Manager are provided below. Some libraries may require purchase of a license.

• Global Arrays — Global Arrays (GA) is a library that allows creation of large matrices spanning multiple nodes. GA is used in a number of Quantum Mechanical packages such as GAMESS-UK and MolPro.
• HDF5 — HDF5 is a Hierarchical Data Format library consisting of a data format specification and a supporting library implementation. HDF5 is designed to address some of the limitations of the older HDF library and to address current and anticipated requirements of modern systems and applications.
• IIPP — The Intel Integrated Performance Primitives (Intel IPP) is an extensive library of multi-core-ready, highly optimized software functions for digital media and data-processing applications. Intel IPP offers thousands of optimized functions covering frequently-used fundamental algorithms. Intel IPP functions are designed to deliver performance beyond what optimized compilers alone can deliver.
• TBB — The Intel Threading Building Blocks (TBB) offer a rich and complete approach to expressing parallelism in a C++ program. It is a library that helps you leverage multi-core processor performance without having to be a threading expert. Threading Building Blocks is not just a threads-replacement library. It represents a higher-level, task-based parallelism that abstracts platform details and threading mechanism for performance and scalability. The Intel TBB requires purchase of a license.
• NetCDF — NetCDF (network Common Data Form) is a set of software libraries and machine-independent data formats that support the creation, access, and sharing of array-oriented scientific data.
• PETSc — PETSc is a suite of data structures and routines for the scalable (parallel) solution of scientific applications modeled by partial differential equations (PDEs). It employs the MPIMessage Passing Interface standard for all message-passing communication.