Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Mar 27, 2025

• Update runners to ubuntu-24.04 from deprecated ubuntu-20.04 label (#51)

  Show description for 2e4c8e4
  

  google-admin

  and

  billnapier

  authored
  2e4c8e4

  Copy full SHA for 2e4c8e4

Commits on Aug 2, 2024

• Add a one workgroup argmax benchmark (#49)

  Show description for 32f2f57
  obtuseangleAZ

  authored
  32f2f57

  Copy full SHA for 32f2f57

Commits on Nov 27, 2023

• Fix subgroup_arithmetic benchmark for flexible subgroup sizes (#46)

  Show description for c9fe9ca
  dneto0

  authored
  c9fe9ca

  Copy full SHA for c9fe9ca

• Fix memory benchmarks for unexpected gl_SubgroupSize (#44)

  Show description for 90866c8
  dneto0

  authored
  90866c8

  Copy full SHA for 90866c8

Commits on Nov 3, 2023

• Add RenderDoc integration (#41)

  Show description for e5af042
  kuhar

  authored
  e5af042

  Copy full SHA for e5af042

Commits on Nov 2, 2023

• Add vector-times-matrix-transposed benchmark (V2) (#40)

  Show description for 3049af9
  

  kuhar

  and

  qedawkins

  authored
  3049af9

  Copy full SHA for 3049af9

Commits on Feb 8, 2023

• [matmul] Add naive mmt benchmark (#33)

  Show description for f3180c7
  kuhar

  authored
  f3180c7

  Copy full SHA for f3180c7

Commits on Feb 5, 2023

• [memory] Copy multiple elements per thread (#32)

  Show description for 70edb25
  kuhar

  authored
  70edb25

  Copy full SHA for 70edb25

Commits on Feb 1, 2023

• [matmul] Tweak innerproduct i8->i32 implementation (#31)

  Show description for e2e343d
  kuhar

  authored
  e2e343d

  Copy full SHA for e2e343d

Commits on Jan 28, 2023

• [matmul] Add basic matmul tiled for inner product (#30)

  Show description for c64b06f
  kuhar

  authored
  c64b06f

  Copy full SHA for c64b06f

Commits on Jan 27, 2023

• Compile shader permutations in parallel (#29)

  Show description for b8218d4
  kuhar

  authored
  b8218d4

  Copy full SHA for b8218d4

Commits on Jan 26, 2023

• [matmul] Clear output buffer between benchmark runs (#28)

  kuhar

  authored
  a3452ec

  Copy full SHA for a3452ec

• Fix i32 matmul output check (#27)

  kuhar

  authored
  5b1be93

  Copy full SHA for 5b1be93

• Add i32 matmul benchmark variant (#26)

  Show description for 555b9cf
  kuhar

  authored
  555b9cf

  Copy full SHA for 555b9cf

Commits on Jan 23, 2023

• Fix bindings mismatch with f16 matmul (#25)

  Show description for 1010cf4
  kuhar

  authored
  1010cf4

  Copy full SHA for 1010cf4

Commits on Jan 20, 2023

• Bump submodules (#24)

  Show description for 892d7ae
  kuhar

  authored
  892d7ae

  Copy full SHA for 892d7ae

• Add i8 tiled matmul benchmark (#23)

  Show description for ecb578d
  kuhar

  authored
  ecb578d

  Copy full SHA for ecb578d

Commits on Jan 19, 2023

• Remove unused CircleCI configuration

  antiagainst

  committed
  b17e3c2

  Copy full SHA for b17e3c2

Commits on Jan 17, 2023

• [CI] Update lint CI action (#22)

  Show description for 5f9a6e4
  kuhar

  authored
  5f9a6e4

  Copy full SHA for 5f9a6e4

• Clean up before introducing i8 benchmarks (#21)

  Show description for 3e2778b
  kuhar

  authored
  3e2778b

  Copy full SHA for 3e2778b

Commits on Apr 1, 2022

• Use explicit constructor to fix GCC compilation issue (#19)

  tpoisonooo

  authored
  51f73fe

  Copy full SHA for 51f73fe

Commits on Jan 22, 2022

• Update Appveyor CI Vulkan SDK to version 1.2.182

  antiagainst

  committed
  4c2cc75

  Copy full SHA for 4c2cc75

• Update Abseil and Google Benchmark

  Show description for 744df66
  antiagainst

  committed
  744df66

  Copy full SHA for 744df66

Commits on Sep 14, 2021

• Replace Travis Build badge with GitHub Action one

  antiagainst

  authored
  d894b06

  Copy full SHA for d894b06

Commits on Sep 13, 2021

• Change matmul benchmark to support both Adreno and Mali Valhall (#11)

  antiagainst

  authored
  e9f0adf

  Copy full SHA for e9f0adf

Commits on Sep 12, 2021

• Use GitHub Actions for Linux and Android build (#12)

  Show description for 7002e5a
  antiagainst

  authored
  7002e5a

  Copy full SHA for 7002e5a

• Update Docker images for CI

  Show description for 83ae188
  antiagainst

  committed
  83ae188

  Copy full SHA for 83ae188

• Change depthwise conv2d benchmark to support Adreno too

  antiagainst

  committed
  e1ef0d9

  Copy full SHA for e1ef0d9

• Change conv2d benchmark to support both Adreno and Mali Valhall

  antiagainst

  committed
  ba31b6a

  Copy full SHA for ba31b6a

Commits on Sep 7, 2021

• Make matmul workgroup X, Y configurable via CMake (#10)

  antiagainst

  authored
  9cf7e42

  Copy full SHA for 9cf7e42

Commits on May 25, 2021

• Add additional hints for finding spirv-as & glslc. (#9)

  Show description for 0680f5a
  kdub

  authored
  0680f5a

  Copy full SHA for 0680f5a

Commits on Mar 13, 2021

• Use proper libvulkan on macOS and export compilation database (#8)

  ergawy

  authored
  5e22593

  Copy full SHA for 5e22593

Commits on Jan 27, 2021

• Add a packed f16 version of convolution

  antiagainst

  committed
  2366b3e

  Copy full SHA for 2366b3e

• Add support for f16 version of convolution

  antiagainst

  committed
  1d79a1e

  Copy full SHA for 1d79a1e

Commits on Jan 21, 2021

• Separate matmul f16 shader and add texture support (#7)

  ThomasRaoux

  authored
  e7adb49

  Copy full SHA for e7adb49

Pagination

Previous

Next