3 BLAS: Basic Linear Algebra Subprograms

The BLAS are a set of well defined basic linear algebra operations ([1], [2], [3]). These operations are subdivided into three groups:

• Level 1: operations acting on vectors only (e.g. dot product)
• Level 2: matrix-vector operations (e.g. matrix-vector multiplication)
• Level 3: matrix-matrix operations (e.g. matrix-matrix multiplication)

Efficient machine-specific implementations of the BLAS are available for many modern high-performance computers. The implementation of higher level linear algebra algorithms on these systems depends critically on the use of the BLAS as building blocks. AMD provides, as part of the ACML, an implementation of the BLAS optimized for performance on AMD64 processors.

For any information relating to the BLAS please refer to the BLAS FAQ:

http://www.netlib.org/blas/faq.html

ACML also includes interfaces to the extensions to Level 1 BLAS known as the sparse BLAS. These routines perform operations on a sparse vector x which is stored in compressed form and a vector y in full storage form. See reference [4] for more information.