All pull requests

Pull requests

• Add support for the s390x architecture

  #946·jamesnohilly opened Jul 29, 2026·

  ·
  1

• Feature detect Emscripten with __EMSCRIPTEN__

  #945·hoodmane opened Jul 24, 2026·

  ·

• Add optional all-dims-small (AND) rule for SUP threshold

  #944·jeffhammond opened Jul 16, 2026·

  ·

• Multithread the memory-bound level-1/2 operations (OpenMP)

  #943·jeffhammond opened Jul 16, 2026·

  ·
  1

• Add optional single-thread cutoff for small level-3 problems

  #942·jeffhammond opened Jul 16, 2026·

  ·

• Prototype: weighted-static partitioning for heterogeneous cores

  #941·jeffhammond opened Jul 16, 2026·

  ·

• Add optional huge-page-backed memory pool

  #940·jeffhammond opened Jul 16, 2026·

  ·
  1

• Add cortexx925 and cortexa725 sub-configs for NVIDIA GB10

  #939·jeffhammond opened Jul 16, 2026·

  ·
  2

• Automatically detect hardware type for zen5 and later

  #927·edwsmyth opened Mar 30, 2026·

  ·
  4

• Amd zen4 support

  #925·harsdave opened Mar 26, 2026·

  ·
  1

• Implement optimized ccopyv and zcopyv kernels for Zen/2/3

  #924·harsdave opened Mar 24, 2026·

  ·
  2

• Add m4sme_p and m4sme_e sub-configurations

  #921·Luislo1 opened Mar 16, 2026·

  ·
  14

• gks: zero-initialize ref_cntx before use

  #883·jerryyiransun opened Aug 26, 2025·

  ·
  6

• Conditionally harden barriers.

  #872·devinamatthews opened Jun 22, 2025·

  ·

• sdgemm micro-kernel for the SG2042 RISC-V processor

  #868·evenieri opened May 6, 2025·

  ·
  2

• Proof-of-concept: speeding up gemm reference kernel

  #863·bartoldeman opened Mar 9, 2025·

  ·
  13

• Add skew-symmetric BLAS operations

  #805·devinamatthews opened Apr 26, 2024·

  ·
  21

• Implement "sandwich product" matrix operations

  #798·devinamatthews opened Feb 16, 2024·

  ·

• Armv8-A Row-major Kernel Improvements

  #698·xrq-phys opened Dec 16, 2022·

  ·
  3

• fix compilation issue for haswell + gcc 7.3.1 (impossible constraints)

  #692·springer13 opened Nov 17, 2022·

  ·
  2

  

• Overhaul error handling in the core framework

  enhancement
  #633·fgvanzee opened May 12, 2022·

  ·
  

• Add support for c_next in the auxinfo_t struct.

  enhancement
  #632·fgvanzee opened May 11, 2022·

  ·
  1

  

• add method to query HW size

  enhancement
  #630·jeffhammond opened May 3, 2022·

  ·
  20

  

• Optional offloading to AMD GPUs

  #626·iotamudelta opened Apr 21, 2022·

  ·
  2

• Create a valid Neoverse N1 target.

  #623·everton1984 opened Mar 14, 2022·

  ·
  11