Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Sep 23, 2025

• perf: Add FP16 GEMM MMUL Reshaped Only Rhs Kernel

  Show description for 531a496
  ArmDude

  committed
  531a496

  Copy full SHA for 531a496

Commits on Apr 25, 2024

• Add update/index/output (m+1)/2d/(m+n) support for CLScatter

  Show description for ada3200
  gunes-arm

  committed
  ada3200

  Copy full SHA for ada3200

Commits on Apr 22, 2024

• Scatter GPU Kernel Implementation for 1D tensors.

  Show description for 7377107
  SuhailMunshi

  committed
  7377107

  Copy full SHA for 7377107

Commits on Oct 31, 2023

• Optimize CL softmax

  Show description for 29254ae
  doviethoa-at-work

  committed
  29254ae

  Copy full SHA for 29254ae

Commits on Sep 28, 2023

• Apply clang-format on repository

  Show description for afd38f0
  felix-johnny

  committed
  afd38f0

  Copy full SHA for afd38f0

Commits on Sep 14, 2023

• Add skeleton of ClMatMulLowpNativeMMULKernel

  Show description for e87fa66
  gunes-arm

  committed
  e87fa66

  Copy full SHA for e87fa66

Commits on Sep 4, 2023

• Remove legacy PostOps code

  Show description for 0d27b2e
  sleepy-hat

  committed
  0d27b2e

  Copy full SHA for 0d27b2e

Commits on Jun 29, 2023

• Implement FP32/16 MatMul Lhs T Rhs T/NT kernel using MMUL extension

  Show description for 00474e9
  gunes-arm

  committed
  00474e9

  Copy full SHA for 00474e9

Commits on Jun 23, 2023

• Implement FP32/FP16 MatMul NT/T kernel using the MMUL extension

  Show description for c952596
  ramelg01

  committed
  c952596

  Copy full SHA for c952596

Commits on Jun 19, 2023

• Implement FP32/FP16 MatMul NT/NT kernel using the MMUL extension

  Show description for a8d8058
  sicong-li-arm

  committed
  a8d8058

  Copy full SHA for a8d8058

Commits on Apr 27, 2023

• Add quantized CL MatMul kernel for LHS NT, RHS T

  Show description for 5e99a3e
  sleepy-hat

  committed
  5e99a3e

  Copy full SHA for 5e99a3e

Commits on Apr 20, 2023

• Implement CL kernel for a native batched matmul Quantized - LHS transposed, RHS transposed

  Show description for 467daef
  ArmDude

  committed
  467daef

  Copy full SHA for 467daef

Commits on Apr 17, 2023

• Add quantized CL MatMul kernels for Lhs NT/T, Rhs NT

  Show description for 9d0c4de
  gunes-arm

  committed
  9d0c4de

  Copy full SHA for 9d0c4de

Commits on Mar 20, 2023

• Implement OpenCL MatMul for Lhs T Rhs T/NT FP32/16

  Show description for 8918b23
  gunes-arm

  committed
  8918b23

  Copy full SHA for 8918b23

Commits on Mar 17, 2023

• Implementation of RSQRT for quantized int8

  Show description for 14d7b53
  ramelg01

  committed
  14d7b53

  Copy full SHA for 14d7b53

• Implement OpenCL MatMul for Lhs NT Rhs T/NT FP32/16

  Show description for 2b6ebfe
  

  ramelg01

  and

  gunes-arm

  committed
  2b6ebfe

  Copy full SHA for 2b6ebfe

Commits on Dec 13, 2022

• Add CLAMP operator to Dynamic Fusion interface

  Show description for 3274172
  sleepy-hat

  committed
  3274172

  Copy full SHA for 3274172

Commits on Dec 9, 2022

• Implement the OpenCL kernel to compute the indirect convolution

  Show description for 76335eb
  gmiodice

  committed
  76335eb

  Copy full SHA for 76335eb

Commits on Nov 25, 2022

• Implement address precalculation for indirect conv2d - OpenCL

  Show description for 5d01681
  gmiodice

  committed
  5d01681

  Copy full SHA for 5d01681

Commits on Nov 14, 2022

• Optimize Transposed Convolution for CL backend (FP32/16)

  Show description for ec0113d
  gunes-arm

  committed
  ec0113d

  Copy full SHA for ec0113d

Commits on Jul 22, 2022

• Add GemmLowp MMUL Reshaped Only Rhs Support for QASYMM8/QASYMM8_SIGNED

  Show description for e572dff
  

  FreddieLiardet

  authored and

  gunes-arm

  committed
  e572dff

  Copy full SHA for e572dff

Commits on Jul 13, 2022

• Add Gemm MMUL Reshaped Only Rhs Support for FP32/FP16

  Show description for 4bfc70e
  gunes-arm

  committed
  4bfc70e

  Copy full SHA for 4bfc70e

Commits on Apr 19, 2022

• Add CLPool3d Int8 Support

  Show description for 5e549fa
  SuhailMunshi

  committed
  5e549fa

  Copy full SHA for 5e549fa

Commits on Apr 14, 2022

• Include missing embedded headers

  Show description for ca364df
  sicong-li-arm

  committed
  ca364df

  Copy full SHA for ca364df

Commits on Mar 31, 2022

• Fix embedded kernel header inclusion for dynamic fusion

  Show description for 892b70a
  giorgio-arenarm

  committed
  892b70a

  Copy full SHA for 892b70a

Commits on Mar 15, 2022

• Implementation of ClPooling3d

  Show description for 3751569
  ramelg01

  committed
  3751569

  Copy full SHA for 3751569

Commits on Feb 9, 2022

• Remove deprecated remap functions.

  Show description for 6863fa0
  AhSinan

  committed
  6863fa0

  Copy full SHA for 6863fa0

Commits on Feb 2, 2022

• Revert "Rework gemm_mm_reshaped_only_rhs_ kernels with new macros"

  Show description for 451c309
  

  Ramy Elgammal

  authored and

  ramelg01

  committed
  451c309

  Copy full SHA for 451c309

Commits on Jan 25, 2022

• Rework gemm_mm_reshaped_only_rhs_ kernels with new macros

  Show description for 10e88a7
  

  gmiodice

  authored and

  SheriZhang

  committed
  10e88a7

  Copy full SHA for 10e88a7

Commits on Dec 13, 2021

• Remove padding from ClDirectConv2dKernel

  Show description for 3012435
  AhSinan

  committed
  3012435

  Copy full SHA for 3012435

Commits on Nov 20, 2021

• Improve start-up timer for GeMM (floating-point):

  Show description for 9cca592
  

  ramelg01

  authored and

  gmiodice

  committed
  9cca592

  Copy full SHA for 9cca592

Commits on Nov 2, 2021

• Add post ops to ClGemmMatrixMultiplyReshapedOnlyRHSKernel and ClGemmMatrixMultiplyNativeKernel Part 3

  Show description for afa1972
  sicong-li-arm

  committed
  afa1972

  Copy full SHA for afa1972

Commits on Oct 28, 2021

• Add experimental PostOp interface to ClGemmMatrixMultiplyReshapedKernel Part 1

  Show description for 1af5416
  sicong-li-arm

  committed
  1af5416

  Copy full SHA for 1af5416

Commits on Oct 18, 2021

• Remove legacy GeMM kernels on OpenCL

  Show description for c9cecc0
  gmiodice

  committed
  c9cecc0

  Copy full SHA for c9cecc0

Commits on Oct 14, 2021

• Implement CLDirectConv3D f32/f16

  Show description for 945ae9e
  giorgio-arenarm

  committed
  945ae9e

  Copy full SHA for 945ae9e