jakpiase dfa95522d3 [CK_TILE] Add support and tests for V6 pipeline in conv fwd (#4357) ...

Added support for conv v6 pipeline in ck tile's convolution forward
kernel. CK Tile v6 pipeline is the equivalent to old ck's V5 pipeline
and should be faster than other pipelines for some cases. This PR also
adds tests inside profiler that's currently inside experimental
directory, so now we should be able to detect regressions easier.

---------

Co-authored-by: Illia Silin <98187287+illsilin@users.noreply.github.com>
Co-authored-by: subhajitdchow <sduttach@amd.com>

2026-02-08 20:57:14 +01:00

..

CMakeLists.txt

Update AMD buffer coherency (#3403)

2025-12-18 10:16:22 +01:00

conv_configs.hpp

[CK_TILE] Add support and tests for V6 pipeline in conv fwd (#4357)

2026-02-08 20:57:14 +01:00

grouped_convolution_backward_data_invoker.hpp

[CK-Tile] move out memory operation from cshuffle epilogue class (#3359)

2026-01-04 03:28:14 -08:00

grouped_convolution_backward_data.cpp

chore(copyright): update copyright header for example directory (#3273)

2025-11-24 18:02:41 -08:00

grouped_convolution_backward_weight_invoker.hpp

[CK-Tile] move out memory operation from cshuffle epilogue class (#3359)

2026-01-04 03:28:14 -08:00

grouped_convolution_backward_weight_two_stage_invoker.hpp

[CK-Tile] move out memory operation from cshuffle epilogue class (#3359)

2026-01-04 03:28:14 -08:00

grouped_convolution_backward_weight_two_stage.cpp

chore(copyright): update copyright header for example directory (#3273)

2025-11-24 18:02:41 -08:00

grouped_convolution_backward_weight.cpp

chore(copyright): update copyright header for example directory (#3273)

2025-11-24 18:02:41 -08:00

grouped_convolution_forward_bias_clamp.cpp

chore(copyright): update copyright header for example directory (#3273)

2025-11-24 18:02:41 -08:00

grouped_convolution_forward_invoker.hpp

[CK-Tile] move out memory operation from cshuffle epilogue class (#3359)

2026-01-04 03:28:14 -08:00

grouped_convolution_forward_large_tensor_invoker.hpp

[CK-Tile] move out memory operation from cshuffle epilogue class (#3359)

2026-01-04 03:28:14 -08:00

grouped_convolution_forward_large_tensor.cpp

chore(copyright): update copyright header for example directory (#3273)

2025-11-24 18:02:41 -08:00

grouped_convolution_forward.cpp

[CK_TILE] Merge multiple fwd convolution groups into a single GEMM batch. (#3136)

2025-12-02 15:23:32 +02:00

grouped_convolution_utils.hpp

[CK-Tile] move out memory operation from cshuffle epilogue class (#3359)

2026-01-04 03:28:14 -08:00

run_grouped_convolution_bwd_data_example.inc

[CK, CK_TILE] Add GPU Reference Implementations for Grouped Convolution (#3216)

2025-12-03 21:14:21 +02:00

run_grouped_convolution_bwd_weight_example.inc

[CK_TILE] Split-K autodeduction (#3351)

2025-12-10 09:30:30 +02:00

run_grouped_convolution_fwd_bias_clamp_example.inc

[CK, CK_TILE] Add GPU Reference Implementations for Grouped Convolution (#3216)

2025-12-03 21:14:21 +02:00

run_grouped_convolution_fwd_example.inc

[CK, CK_TILE] Add GPU Reference Implementations for Grouped Convolution (#3216)

2025-12-03 21:14:21 +02:00