AMD Optimizing CPU Libraries (AOCL)

NEW! AOCL 5.0 is now available, October 10, 2024

AOCL is a set of numerical libraries optimized for AMD processors based on the AMD “Zen” core architecture and generations. Supported processor families are AMD EPYC™, AMD Ryzen™, and AMD Ryzen™ Threadripper™ processors. The tuned implementations of industry-standard math libraries enable rapid development of scientific and high-performance computing applications.

Libraries

• AOCL-Data Analytics (New!)
• AOCL-BLAS
• AOCL-LAPACK
• AOCL-Compression
• AOCL-Cryptography
• AOCL-FFTW (Fastest Fourier Transform in the West)
• AOCL-LibM (AMD Math Library)
• AOCL-LibMem
• AOCL-RNG (AMD Random Number Generator Library)
• AOCL-SecureRNG (Secure RNG Library)
• AOCL-ScaLAPACK
• AOCL-Sparse
• AOCL-Utils

What’s new in AOCL 5.0 (October 10, 2024)

• AOCL-Data Analytics 
  • A new library providing optimized building blocks for data analysis and classical machine learning.
• AOCL-BLAS
  • Turin specific tuning for the following APIs:
    • D/ZGEMM, DTRSM, and DNRM2
  • AVX512 improvements for the following APIs:
    • ZGEMV, D/ZAXPYF, D/ZDOTXF, ZDOTV, C/ZSCALV, DNRM2, S/D/ZCOPY, S/D/C/ZAXPBYV, DTRSV, DGEMMT, D/ZTRSM, and D/ZGEMM
  • Improvements to AOCL_ENABLE_INSTRUCTIONS functionality.
  • Additional APIs and Post-Ops support in addition to the improved performance for the existing APIs in aocl_gemm add-on.
• AOCL-Compression
  •  Single-threaded performance improvements for ZLIB and BZIP2.
  • Multi-threaded support for LZ4HC and performance improvements for LZ4, ZSTD, ZLIB, and Snappy.
  • AOCL_ZLIB_QUICK_MODE option for faster compression in ZLIB and AOCL_DECOMPRESS_FAST option for faster decompression in ZSTD and LZ4.
• AOCL-Cryptography
  •  OpenSSL Provider support and fixes for Cipher, CMAC, Poly1305, and RSA-2048 algorithms.
  • Features added:
    • Chacha20-Poly1305 Cipher
    • RSA PKCS Encrypt/Decrypt, PKCS Sign and Verify, and PSS Sign and Verify
    • SHA3 Shake Squeeze
  •  Performance improvements
    
    • OAEP Encrypt Decrypt
    • HMAC, CMAC, and Poly1305
    • Chacha20 stream cipher
    • CFB, CBC, and GCM
    • SHA2 and SHA3
• AOCL-FFTW
  • Support for using the Wisdom feature by default under the –enable-amd-app-opt option
• AOCL-LAPACK
  • Improvements to SVD, Factorization, Inverse, and Least square Solver APIs.
  • Option to set specific ISA code path at runtime through a new environment variable.
  • Sphinx-based AOCL-LAPACK API documentation.
  • pkgconfig support on linux with cmake builds.
• AOCL-LibM
  • Added new vector APIs – powx and linearfrac.
  • Added vector variants of sincos, sqrt, and fabs.
  • Added vector array variants of pow, exp, exp2, exp10, expm1, log, log2, log10, log1p, cbrt, cos, and sin.
  • Debuting a beta version of CMAKE build system for AOCL-LibM on Linux
    
• AOCL-LibMem
  • Added support for new functions:
    • memchr, strcat, strstr
  • Improved AVX2 performance for smaller sizes.
• AOCL-RNG
  • AVX2 support for double precision MRG32K3A Generator.
  • Added Sobol as new generator of type QRNG.
  • Dynamic dispatch feature enhancements.
• AOCL-ScaLAPACK
  • Ability to build ScaLAPACK library without building a test framework.
  • Tracing and Logging feature enabled now for all APIs including single precision.
• AOCL-Sparse
  • Incorporated new APIs:
    • Level 3: sp2md, spmmd, syrk, syrkd, sypr, syprd
    • Sparse preconditioner: sorv
  • Support for:
    • Symmetric and Hermitian matrices in csrmm API
    • Strided dense vector in TRSM and TRSV
    • Symmetric Gauss Seidel
  • Performance improvements:
    •  Level 1: dot, gthr, sctr, roti
    • Level 2: SpMV for complex general matrices
  • Multi-Threading support
    • Level 3: Sp2M, SpMM, SpAdd
  • Benchmarking framework improvements:
    
    • Enhanced statistics, support for new APIs , random matrix generation capability (Hermitian, diagonally dominant) APIs to support HPCG
• AOCL-Utils
  
  • New APIs to get Instruction Set Architecture (ISA) details
  • New API to check CPU features
  • Thread pinning
  • Support for Linux and Windows platforms
• AOCL-Enabled MUMPS Library
  
  • Building for Windows 10 systems: CMake-based build system on AOCL-enabled MUMPS sparse solver library; refer to GitHub MUMPS-build (https://github.com/amd/mumps-build) for more information.
  • Mumps 5.5.1 support with 5.5.0 and 5.4.1 backward compatibility.

Resources and Technical Support

Documentation

AOCL User Guide

AOCL Tuning Guide

AOCL Release Notes

For prior versions of AOCL, refer to AOCL Archive.

Support

For support options, refer to Technical Support.

AMD Community

For moderated forums, refer to the AMD community.

Download with End User License Agreement