BLIS

BLIS is a portable software framework for instantiating high-performance BLAS-like dense linear algebra libraries. The framework was designed to isolate essential kernels of computation that, when optimized, immediately enable optimized implementations of most of its commonly used and computationally intensive operations. Some of these essential kernels have been optimized for the AMD EPYCTM processor family. The optimizations are done for single and double precision routines.

Highlights of AMD BLIS 0.9 Beta

  • Performance of the small matrix GEMM (general matrix-matrix multiplication) has been improved for single thread, non-transpose cases. This avoids the overhead of packing operation of BLIS at small matrix sizes
  • Optimized TRSM routines for single & double precision
  • The fused kernels DOTXF and AXPYF are optimized for faster  GEMV (matrix vector multiplications)
  • Other optimized BLIS level 1 routines include DOT, DOTV, AXPY, NRM2, AMAX and SCALV

The package containing AMD optimized BLIS Library binaries, examples and documentation are available in the Downloads section below.

Source code is available on GitHub page.

libFLAME

libFLAME is a portable library for dense matrix computations, providing much of the functionality present in LAPACK. It includes a compatibility layer, FLAPACK, which includes complete LAPACK implementation. The library provides scientific and numerical computing communities with a modern, high-performance dense linear algebra library that is extensible, easy to use, and available under an open source license. libFLAME is a C-only implementation and does not depend on any external FORTRAN libraries including LAPACK. There is an optional backward compatibility layer, lapack2flame that maps LAPACK routine invocations to their corresponding native C implementations in libFLAME. This allows legacy applications to start taking advantage of libFLAME with virtually no changes to their source code.

In combination with AMD optimized BLIS library, libFLAME enables running high performing LAPACK functionalities on AMD platform. The performance of libFLAME on AMD platforms can be improved by just linking with the AMD optimized BLIS.

Highlights of AMD libFLAME 0.9 Beta

  • Single Core Optimized BLIS & libFLAME libraries for AMD EPYCTM Family processors
  • Optimized libFLAME when used with optimized BLIS as BLAS library.

The package containing libFLAME binaries, examples and documentation are available in the Downloads section below.

Source code is available on GitHub page.

Download:

File Name Version Size Launch Date OS Bitness Description

File Name

Version

0.9 Beta

Size

1,766 KB

Launch Date

06/20/2017

OS

Linux

Bitness

64

Description

BLIS Library binary package with example binary

File Name

Version

0.9 Beta

Size

4,465 KB

Launch Date

06/20/2017

OS

Linux

Bitness

64

Description

libFLAME Library binary package with example binary