* formatted
* formatted
* formatting
* formatting
* formatting
* [CK TILE GEMM] Refactor block_scale_gemm examples
- Split cpp file to reduce building time
- Support multiple GemmConfig
* [CK TILE GEMM] Refactor block_scale_gemm examples
- Update Readme
* enable prefill shapes
* [CK TILE GEMM] Refactor block_scale_gemm examples
- Add support for rowcol and tensor GEMM operations
* [CK TILE GEMM] Refactor block_scale_gemm examples
- Update README
* adding preshuffle quant as new parameter and its associated new files
* remove debugging statements
* adding test
* enable preshuffle quant with permuteN
* updating readme and correcponding gemmconfigs
* updating cmake file
* fixing CI failures for grouped quant gemm
* debugging permuteN
* debugging
* debugging PermuteN
* initial commit
* resolving merge conflicts
* adding test cases
* initial commit with prints
* debugging
* fine-grained working
* debugging medium grained
* fixing the tile window
* formatting
* enabling prefill shapes
* working prefill shapes
* formatted
* clean up
* code cleanup
* bug fix after merging with develop
* G128 working for both prefill and decode shapes for preshufflequant
* clean up after merging with develop
* fixing group 64 for decode shapes
* non preshufflequant working for group size 128
* enable preshuffleb and preshufflequant with variour group sizes
* reduce build time by splitting example into diff datatype files
* Adding tests for preshuffleQuant
* address review comment
* fix for gfx1201
* compile time fix for gfx1201
* clang formatted
---------
Co-authored-by: Cong Ma <congma13@amd.com>
Co-authored-by: Thomas Ning <Thomas.Ning@amd.com>
Co-authored-by: Agarwal <khuagarw@ctr2-alola-login-03.amd.com>
* CMakeLists.txt hack for Windows.
* Add Windows build instructions.
* Fix type issue with variadic min function.
* Use std::common_type to fix the variadic min/max functions.
* Enable CPU guard compilation on Windows.
* Suppress warnings related to std::getenv on Windows platform.
* Git ignore the output directory on Windows platform.
* Powershell script for running tests and generating reports.
* Improve test logging.
* Disable non-conv tests.
* Fix Debug build on Windows.
* More debug build changes.
* Update Windows build instructions.
* Enable all tests.
* Test fixes.
* Suppress not found linker options warning.
* Update unsigned long literals and format specifiers to work correctly in Windows
* Fix conv 3D bwd weight bilinear tests on Windows.
* Revert changes on .gitignore.
* Clean-up CMake project file for Windows builds.
* clang-format
* Fix definition of CMAKE_PREFIX_PATH on both Linux and Windows platforms.
* Fix building examples on Windows.
* Update Readme.
* Remove the suppression of the deprecated warnings.
* Remove Windows specific min/max implementations from CK Tile math core.
* Remove unnecessary no-op on Windows.
---------
Co-authored-by: User <user@example.com>
Co-authored-by: Ville Pietilä <none>
Co-authored-by: John Afaganis <john.afaganis@amd.com>
Co-authored-by: Ville Pietilä <>
* Refactor GPU verification kernel to gather erorr stats on GPU
* Check if result is all zero
* non-negative error count doesn't need custom Atomics
* Remove unnecessary AtomicMaxFloat function
* Simpler warp reduction, remove passed flag
* Move verification header to include
* Fix header path in test
* Fix block reduction loop
- Add support for direct store in epilogue instead of cshuffle
- Add padding support for wave transfer without transpose
- Add wave transfer with interleaved layout to support direct store
- Enable new functionalities on GEMMs
- Add optional new functionality support for grouped convolution fwd
- Add some fast instances for grouped convolution fwd with new functionalities (proper tuning needed)
* memory op changes
* memory op changes
* Fixing TILE_ENGINE_BASIC in Tile Engine
* Removing gfx90a from Tile Engine Run
* [CK TILE ENGINE] increasing ci configs for BASIC case
* Setting RUN_TILE_ENGINE_BASIC_TESTS to ON by default
---------
Co-authored-by: Max Podkorytov <4273004+tenpercent@users.noreply.github.com>
* solve compiler issue
* solve the gfx950 mfma shuffle regression
* refactor jenkinsfile to handle arch name better
* [CK TILE] set divisor to count of thread along k dimension
* fix the compiler error
* solve degradation
* Finish the multiplies fix
* fix the scales
* solve compilation error
* solve the composes
* solve the error of tile sweeper
* fix the test and example
* fix for gfx950
---------
Co-authored-by: Max Podkorytov <4273004+tenpercent@users.noreply.github.com>
Co-authored-by: illsilin_amdeng <Illia.Silin@amd.com>
Co-authored-by: Cong Ma <congma13@amd.com>
* Add placeholder test.
* Initial conv bwd weight factory.
* Conv builder test refactoring.
* Add missing pieces to bwd weight factory.
* Improve compile time erros message when no matching factory is found.
* Use amcro to ensure automatic macthing between concepts are their string representations.
* Improve compile time diagnostics.
* Small improvements.
* Improve missing member/wrong type compile-time errors.
* Improve compile time diagnostics.
* Concept bug fixes.
* Remove debug assert.
* Update algorithm signature diagnostics.
* Factory bug fixes.
* First functional version of bwd weight conv factory.
* Refactor handing of GEMM-K batch template parameter in conv bwd weight factory.
* Concept improvements.
* Improve concept diagnostics.
* Introduve a common size type for concepts.
* Update compiletime diagnostics to use the size type.
* Update conv specialization enum.
* Fix fwd conv builder tests.
* Fix smoke tests.
* Separate bwd weigth and bwd data tests into separate targets.
* Clean-up CK Tile builder tests.
* Add bwd weight XDL CShuffle V3 factory.
* Build conv bwd weigth v3 instances successfully.
* Add instance traits for DeviceGroupedConvBwdWeight_Xdl_CShuffleV3.
* Test fix.
* Add instance traits for bwd weight algorithms.
* Add unit tests for instance strings.
* Build new instance traits unit tests but exclude WMMA for now.
* Added factory for DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle.
* Conv bwd weight DL factory.
* Final implementation for bwd weight DL factory.
* Add test for creating DeviceGroupedConvBwdWeightMultipleD_Xdl_CShuffle instance.
* Add factory for DeviceGroupedConvBwdWeightMultipleD_Xdl_CShuffle
* Treat ref algorithm the same way as real algorithms in the dispatcher.
* Refactor large tensor support and WMMA configuration.
* Add factory and tests for DeviceGroupedConvBwdWeight_Wmma_CShuffleV3.
* Update Readme.
* Fix WMMA bwd weight tests.
* Added factory and tests for DeviceGroupedConvBwdWeightTwoStage_Wmma_CShuffleV3.
* Factory and tests for DeviceGroupedConvBwdWeight_Wmma_CShuffle.
* Dispatching for DeviceGroupedConvBwdWeightMultipleD_Wmma_CShuffle.
* Add factory for DeviceGroupedConvBwdWeightMultipleD_Wmma_CShuffleV3
* Fix DeviceGroupedConvBwdWeightMultipleD_Wmma_CShuffleV3 factory and compute types for input and output tensor in bwd weigth convs.
* Fix fwd factories after refactoring.
* clang-format
* Move compile-time diagnostics to a separate branch.
* Fix ref algorithm dispatching.
* Fix smoke tests.
* clang-format
* Fix factory for regular WMMA conv bwd weight.
* Clarify builder Readme.
* Remove obsolete test file.
* Fix test after merge.
* clang-format
* Remove the C++26 extensions.
* Unify conv elementwise ops and layout definitions for fwd and bwd directions.
* Remove old layout and elementwise ops.
* Unify handling of conv tensor types between fwd and bwd directions.
* Unify block transfer for fwd and bwd directions. Rename ThreadSliceDim to ThreadClusterRank.
* Make BlockTransferDescriptor concept parametrized. Introduce a common TileTransferParameters concept for conv algorithms.
* clang-format
---------
Co-authored-by: Ville Pietilä <>
* feat: grouped gemm tile loop support for RDNA4
* fix: removed extra parameter from grouped gemm example instance
* fix: FP8 check incorrectly enabling FP8 on RDNA3
This commit addresses numerical stability issues in the BlockFmhaPipelineQRKSVS pipeline when bias has -inf masking values:
1. Explicitly handle the case where the accumulated exponential sum (l) is zero. In this case, the LSE is now correctly set to negative infinity, preventing log(0) errors.
2. Extend the zero-check protection in the normalization step to cover the ELEMENTWISE_BIAS case, preventing potential division by zero.
- Enable page_size=1 support in batch prefill codegen (linear layout only).
- Implement per-token page lookup in `kv_offset_array_transform` for page_size=1 to handle 3D input tensors correctly.
- Relax `kPageBlockSize` alignment assertion for the page_size=1 case.
* refactor: remove Default scheduler implementation as it not used anymore
* refactor: remove dead code from gemm universal kernel
* chore: add descriptive comments about amd intrinsic hardware sync instructions
* fix: label existing memory pipeline for aquant as intrawave
This adds some utilities to automatically generate UniqueInputs,
UniqueOutputs, alloc_inputs, alloc_outputs, and validate, based
on a Inputs::reflect() and Outputs::reflect().
* nt on fp8 blockscale
* some improve and tests needs to be fixed
* update
* fix format
* revert useless change
* revert any change in amd_buffer_coherence
* WIP
* Add Unit tests for the Multi Reduction Kernel
* clang format
* Rename multiblock to threadwise
* Multiblock WIP
* Fix multi reduce multi block unit tests
* Multi Reduce Tile Engine: WIP
* refactoring + try addressing precision error
* Fix multiops examples
* Cleanup
* Clean up tile engine's reduce op
* Update changelog
* Fix remod/clang
* Fix dates
* Fix documentation & missing file
* Fix comments
* Use the update_tile api in the multi-block kernel
* Unify threadwise/multiblock into a single kernel + default multiblock output to float in tests
* Add TileParitioner
* Cleanup
* Add warning when no data to process, in the example
* Refactoring Reduce kernel Tile Partioner + cleanup
* Move the tile partioner to its own file
* Add missing includes
* Fix copyright header with update_amd_copyright_headers.py
* Fix change of interface in Reduce2dProblem
---------
Co-authored-by: Damien Lejeune <damien.lejeune@amd.com>
Co-authored-by: Adam Osewski <19374865+aosewski@users.noreply.github.com>
* ck-builder: make toString to_string
We are using snake case for CK-Builder
* ck-builder: add debug.hpp with tensor descriptor printing function
This adds some initial functionality to debug.hpp, a header which will
be used to house some debug utilities.
* ck-builder: abstract nd-iteration
Abstracting this makes it easier to test, clearer, and allows us to
use it elsewhere (such as in debug.hpp soon)
* ck-builder: tensor printing
* ck-builder: rename INT32 to I32
This makes it more in line with the other data type definitions.
* Take split_k into account when checking 2GB tensor limit.
* Revert "Take split_k into account when checking 2GB tensor limit."
This reverts commit adf35c91be.
* Optimize grouped conv bwd wei split_k off calc
(cherry picked from commit 6f61dd56c5)
* Update gridwise_gemm_xdl_cshuffle_conv_v3.hpp
(cherry picked from commit b33877c10f)
* Fix tensor descriptors and stride calculations
* Don't miss half of the elements
* Fix buffer size calculations
* Disable hack if stride not divisible by k_batch
* Clean up comments
* Disallow hack in non-contiguous edge cases
* Index -> Dim
* Fix broken test
* Refactor applicability checks into separate function
* fix missed variable name
* Fix variable name in info print
* update V3 2GB check
* No more regression, use templates instead
* Code deduplication
* Regression fix for cshuffle
* arch-guarded atomic_add implementations for gfx11
* Similar for half(4|8)_t as well
* Only use both offset hacks at the same time
* Revert "arch-guarded atomic_add implementations for gfx11"
This reverts commit 3883fe6935.
This reverts commit 5311ec608d.
* Reapply "arch-guarded atomic_add implementations for gfx11"
This reverts commit 1972adeddc.
* Only remove float4 atomic_add
* Refactor to single flag
* Consolidate template parameters
* Consolidate flag in transformers
---------
Co-authored-by: Bartlomiej Kocot <barkocot@amd.com>
* Implement grouped gemm fastgelu for RDNA4
* chore: some cleanup and minor inconsistencies in grouped gemm profiler
* chore: clarified logic and reporting of supported instance warnings
Our concept-base conversions are fragile and too complex. We want to refactor to straightforward functions
for each intance trace class template. This change adds unit test coverage to make that refactoring safer.
Refactor function amd_buffer_load_invalid_element_return_zero to avoid
the inefficient ASM code generated by compiler.
Compiler generates suboptimal assembly for ternary operator, causing excessive VGPR usage
Tested compilers:
- Rocm 7.0.1
- Rocm 7.1.1
Co-authored-by: Thomas Ning <Thomas.Ning@amd.com>
* update grouped_gemm blockwise kernel
* update config
* update kernel
* update examples
* remove test code for now
* sync test files with origin/develop
* update example
* fix code lint
* fix code-lint
* update test code
* run clang format
* run pre-commit
* update api
* ck-builder: explicitly delete forward declarations
Before, these functions were seen as a forward declaration for an existing function.
If no actual implementation overload could be found, these would be selected and
a linker error or warning would be generated. By marking these functions as explicitly
deleted, they incorrect invocations are generated as compile error instead.
* ck-builder: ckt::run plumbing for reference conv
This implements the ckt::run plumbing for the reference convolution
implementation and sets up the first complete end-to-end test.
* ck-builder: make validation system check for all-zeros
When both the actual and reference output are both all zero bits,
there is probably something wrong in the test framework.
* ck-builder: proper implementation+tests for TensorDescriptor::is_packed
* ck-builder: fix typos
* [CK_TILE] unify double and single lds implementation (#108)
Unify LDS buffer management API for single and double buffering modes
This change consolidates the Local Data Store (LDS) buffer management by:
Merging single and double LDS buffer APIs into a unified interface
Implementing ping-pong address calculation in pipeline when double LDS is enabled
Computing pong buffer addresses dynamically using base address offsets
---------
Co-authored-by: joye <joye@amd.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* update wp_pipeline
* fix a c++17 issue
* update for ci errors
* fix ci issues
* include a header to fix ci errors
* fix some rebase issues
* update with rebase
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
This pull request builds on #3267 by proving the "validation" infrastructure, the means to compare a set of `Outputs`.
The design of the validation infrastructure is relatively straight forward:
- Each SIGNATURE should come with a `validate()` implementation, which should be implemented in a similar way that the other functions/types from `testing.hpp` are implemented.
- `validate()` returns a `ValidationReport`, which is a structure that keeps all relevant information about comparing the tensors from two `Outputs`. Note that crucially, `validate()` should not do any reporting by itself. Rather, glue logic should be implemented by the user to turn `ValidationReport` into a relevant error message.
- You can see this clue code for CK-Builder itself in `testing_utils.hpp`, its `MatchesReference()`. This functionality is relatively barebones right now, it will be expanded upon in a different PR to keep the scope of this one down.
The comparison is done on the GPU (using an atomic for now), to keep tests relatively quick. Some notable items from this PR:
- To help compare the tensors and with writing tests, I've written a generic function `tensor_foreach` which invokes a callback on every element of a tensor.
- For that it was useful that the `TensorDescriptor` has a rank which is known at compile-time, so I've changed the implementation of `TensorDescriptor` for that. I felt like it was a better approach than keeping it dynamic, for multiple reasons:
- This is C++ and we should use static typing where possible and useful. This way, we don't have to implement runtime assertions about the tensor rank.
- We know already know the rank of tensors statically, as it can be derived from the SIGNATURE.
- It simpifies the implementation of `tensor_foreach` and other comparison code.
- There are a lot of new tests for validating the validation implementation, validating validation validation tests (Only 3 recursive levels though...). For a few of those functions, I felt like it would be useful to expose them to the user.
- Doc comments everywhere.
* add page_block_size parameter
* add is_sglang_layout to parameters
* add kv_offset_array_transform to batch async for page size 16
* add kv_last_page_lens to kernel
* change kv layout to [num_total_pages, page_block_size, hdim]
* format
* - enable codegen of batch_prefill kernels
- create new problem struct BlockFmhaBatchPrefillPipelineProblem for
batch prefill kernels
- generate different page sizes of batch prefill kernels (1, 16)
* 1. fix wrong calculation of page id in kv_offset_array_transform in gfx950
2. support page size 1024
* fix python format
* change kv cache layout to [num_blocks, num_kv_heads, head_size/x,
block_size, x] and [num_blocks, num_kv_heads, block_size/X, head_size, X]
* 1. Introduced `kVectorSize` in BlockFmhaBatchPrefillPipelineProblem instead of using hardcode values
2. Makes batch prefill kernel traits structures inherent from fmha fwd
traits
3. Add some static check for Page size, vector size, hdim, ..., etc.
* [Refactor] Replace is_sglang_layout with Enums for KV cache configuration
Refactored `fmha_batch_prefill` to use `BlockAttentionKVCacheMemoryLayoutEnum` (VECTORIZED/LINEAR) and `BlockAttentionKVCacheLookupTableEnum` (SGLANG_1D/VLLM_2D) instead of a single
boolean.
**Changes:**
* Added Enum definitions in `block_attention_kvcache_layout_enum.hpp`.
* Updated Kernel, Pipeline, and Traits to template on these Enums.
* Implemented `kv_offset_array_transform` logic based on `kKVMemoryLayout`.
* Refactored `PageBlockTableKargs` to adapt to `kKVLookupTable`.
* Updated CodeGen scripts to support new parameters.
This decouples memory layout from the paging mechanism, enabling flexible KV cache configurations.
* 1. remove batch prefill pipeline with sk_pad=false
2. correct some comments
3. add static assert to make sure v offsets is in same page within a tile.
* fix vgpr spill count
* remove unnecessary t2s functions
* add fp8 support for receipt 200 and 600 in fmha_bath_prefill.py
* support linear kv cache layout
* Remove block_table_ptr from fwd_batch_prefill_args. Instead, reuse
kv_page_indices as a pointer of the lookup table.
* 1. merge multiple transforms into single transform.
2. add static check to make sure vlayout is row-major.
* move FmhaFwdCommonKargs::seqlen_k_ptr to VllmPageTableKargs.
* update changelog
---------
Co-authored-by: ltqin <letaoqin@amd.com>
Co-authored-by: PoYen, Chen <PoYen.Chen@amd.com>
* initial poc
* factor out common parts in operator()
* cv4
* rest of the universal gemm pipelines
* fix test
* remove boilerplate from tile engine
* fix example
* fix example
* format
* fix tests build for gemm
* remove base pipeline codegen from gemm instance builder
* unify v3 logic with the rest of universal gemm pipelines
* fix build for multi abd test
* fix test gemm multi d
* fix build for weight preshuffle
* fix grouped gemm test
* fix grouped gemm multi d test
* fix grouped gemm preshuffle
* fix grouped gemm example except for quant
* fix gemm preshuffle
* fix splitk 2 stage example
* fix batched gemm example
* fix multid example
* fix multiabd example
* fix batched gemm test
* fixup
* fix examples build
* fix grouped gemm test build
* fix smoke builder
* hacky poc
* fix tile engine
* kill the lambda
* maybe fix test build
* more fixes
* clang-format
* save temp
* clang-format
* mostly fix examples
* clang-format
* remove dead code
* more cleanup
* fix fmha bwd build (default epilogue set/add appears to be broken)
* fix default epilogue tests but not correctness
* clang-format
* fix bquant
* clang-format
* cleanup dead code
* rearrange make windows for readability
* restore changes to IsSupportedArgument
* fix smoke-builder
* clang-format
* fixup rename class
* build fixes
* clang-format
* fix builder
* fixup
* remove set from builder tests
* fix test
* clang-format
* re-refactor the kernels
* clang-format
* fix header license
* remove memory operation from conv bwd test
* clang-format
* clang-format example,include
* clang-format test
* build fixes
* clang-format
* solve compilation error
* fix the CI
* solve compilation error
* clang format
* solve merge conflict
* solve merge conflict
* solve the gfx11 error
* solve test error
* moar build fixes
* remove AtomicAddRequiresKBatchGreaterThanOne test since the property is removed from the kernel scope
---------
Co-authored-by: Thomas Ning <Thomas.Ning@amd.com>
Introduces a new Python toolset in script/analyze_build/ for analyzing
Clang -ftime-trace JSON output to identify compilation bottlenecks and
optimize C++ metaprogramming build times.
Key features:
- Fast parallel processing of trace json files using all CPU cores (> 100 files/sec)
- Simple, cache-free architecture for consistent performance
- Comprehensive analysis of template instantiations and event types
- Command-line tools and Jupyter notebook support
- Automatic orjson detection for JSON parsing speedup
Components:
- trace_analysis/: Core library (models, parser, transformer)
- examples/: CLI tools for single-file and directory analysis
- notebooks/: Comprehensive Jupyter notebook with analysis patterns
- Detailed README with usage examples and performance data
Also adds ruff configuration to pyproject.toml to ignore E402 (module
level import not at top of file) for Jupyter notebooks, which commonly
have imports after markdown cells.
This toolset addresses the critical problem of long build times in CK's
C++17 metaprogramming codebase by treating -ftime-trace as a big data
problem, using pandas and modern analysis tools to understand compilation
patterns and measure improvement opportunities.
