Welcome to ftp.nluug.nl Current directory: /os/Linux/distr/salix/sbo/15.0/libraries/libxsmm/ |
|
Contents of README:LIBXSMM is a library for specialized dense and sparse matrix operations as well as for deep learning primitives such as small convolutions. The library is targeting Intel Architecture with Intel SSE, Intel AVX, Intel AVX2, Intel AVX512 (with VNNI and Bfloat16), and Intel AMX (Advanced Matrix Extensions) supported by future Intel processor code-named Sapphire Rapids. Code generation is mainly based on Just-In-Time (JIT) code specialization for compiler-independent performance (matrix multiplications, matrix transpose/copy, sparse functionality, and deep learning). LIBXSMM is suitable for "build once and deploy everywhere", i.e., no special target flags are needed to exploit the available performance. Supported GEMM datatypes are: FP64, FP32, bfloat16, int16, and int8. NOTE: the library does not support 32-bit architecture (64-bit only) |
Name Last modified Size
Parent Directory - README 20-Oct-2022 18:35 862 libxsmm.SlackBuild 20-Oct-2022 18:35 3.4K libxsmm.info 20-Oct-2022 18:35 322 slack-desc 20-Oct-2022 18:35 1.1K
NLUUG - Open Systems. Open Standards
Become a member
and get discounts on conferences and more, see the NLUUG website!