AMD’s Core Math Library (ACML 6 GA) is now available as a production SDK, as announced in this blog post. This is the first version of ACML that supports heterogeneous computing, leveraging the open source clMathLibraries projects as a compute backend “offloading significant computation to OpenCL devices”.

ACML 6 introduces:

  • A subset of the FFTW interfaces for heterogeneous compute
  • BLAS level 3 and two level 2 (GEMV/SYMV) routines supported for heterogeneous compute
  • A scripting language to tune the heterogeneous behaviour

ACML is a tuned math library designed for high performance on AMD64 machines, including Interlagos, Opteron and Athlon 64, and includes both 32-bit and 64-bit library versions. Different versions are available for Linux and Windows operating systems.

ACML consists of the following main components:

  • A full implementation of Level 1, 2 and 3 Basic Linear Algebra Subroutines (BLAS), with key routines optimized for high performance on AMD Opteron™ processors
  • A full suite of Linear Algebra (LAPACK) routines. As well as taking advantage of the highly-tuned BLAS kernels, a key set of LAPACK routines has been further optimized to achieve considerably higher performance than standard LAPACK implementations
  • A comprehensive suite of Fast Fourier Transforms (FFTs) in both single-, double-, single-complex and double-complex data types; and 4) Random Number Generators in both single- and double-precision

Full details at: http://developer.amd.com/