* Implement argument passing to element-wise functions for fwd convolution
* Add files for fwd + bias + clamp example
* Implement Bias
* Implement Clamp
* Elementwise function composition
* Composition unit test
* Implement fwd + bias + clamp example
* Simplify argument passing and composition
* elfunc -> bias_and_clamp
* Rename function to specify example
* Move element-wise function instantiation to kernel
* Make bias a runtime tensor
* No ugly namespace aliasing
* Initialize element-wise function on host
* Remove function initialization helper, simplify Compose initialization
* Remove unintended LSP compatibility patch
* Clean up includes and unused code
* Switch names in cshuffle epilogue
* Move CDElementwise to conv traits
* Re-add required include
* Initialize bias in same way as other tensors
* Better type specification for ds pointer
* Disable 1D convolution
* Add warning for non-group-constant bias
* Persistent Stream-K Kernel Implementation
This change implements an operator() function in the
reboot::StreamKKernel class that is enabled when the Persistent flag is
set to true. In this case, the data-parallel portion and the Stream-K
portion of the kernel are fully persistent.
The changes were made in the reboot namespace. A future PR will remove
the old Stream-K kernel class and remove the reboot namespace.
* Unit Tests for Persistent Stream-K Kernel
This change contains the inital test suite for the Persitent Stream-K
Kernel. The files contain "reboot" in the name; a future PR will remove
tests for the old Stream-K Kernel and remove the "reboot" naming.
A future commit will add tests for the non-persistent kernel.
Also added estimate_num_wgs_per_tile to the StreamKTilePartitionerBase
class. This allows us to estimate the number of accumulations done per
macro tile in C to use during validation when computing relative and
absolute tolerance.
* Adding implementation for the Non-Persistent Stream-K kernel
This code is adding the operator() function for the Non-Persistent Stream-K
kernel. Persistency of the kernel is determined through a template argument.
The Non-Persistent kernel will allocate additional workgroups for the data
parallel section, leading to a different structure for processing the data
parallel and Stream-K sections.
There has been an addition to the TilePartitioner to get access to the whether
Persistent has been set to true or false in the StreamKKernel.
* Adding in the tests for the Non-Persistent Stream-K kernel
* Refactor Stream-K Reboot Unit Tests
This commit makes the following changes:
- Update test cases to determine M, N, and K based on the number of CUs.
This ensures that each test case is one of Edge Case, SK Only, DP
Only, or DP + 2 Tile SK regardless of the architecture.
- Since the DP + 2 Tile SK test case takes long to run, this change
moves this case into a separate .inc file and labels it as an extended
test.
- Since the extended test takes > 30 seconds to run, this test is added
to the list of regression tests.
* Fix spelling errors in comments for test cases
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Changes based on review
Removed const volatile for typenames
Set up alias for is_tuple_t
Naming changes for clarity: GemmCommon -> BaseGemm
Moved std::enable_if_t out of template parameters and changed to a return type for operator()
Added constructor for StreamKKernelArgs to clarify UniversalGemm inheritance
---------
Co-authored-by: Emily Martins <emily.martins@amd.com>
Co-authored-by: Christopher Millette <63608002+cgmillette@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Add name member to unary elementwise ops.
* Update elementwise_op_name to check for name attribute.
* Require that the layout is derived from BaseTensorLayout struct.
* add tensorwise quant in grouped gemm
* fix example issue
* update test cases
* format codes
* clang format
* use GTEST_FAIL
* fix a bug in test_grouped_gemm_util
* skip test when use wmma on grouped_quant kernel
* change cmake
* change code based on comments
---------
Co-authored-by: ThomasNing <thomas.ning@amd.com>
* Introduce tree reduction for BlockReduce2dCrossWarpSync
* Rename original impl to BlockReduce2dLinearCrossWarpSync
* Replace warp_size with get_warp_size()
---------
Co-authored-by: Illia Silin <98187287+illsilin@users.noreply.github.com>
Prior to this change, the number of accumulations passed into
calculate_rtol_atol was 1. That said, in most cases, this is not correct
when there are multiple workgroups contributing to the same macro tile
in C.
This change ensures uses the function estimate_num_wgs_per_tile, which
was extracted into a common file and generalized, to estimate the number
of workgroups per macro tile. This estimate is passed into
calculate_rtol_atol to ensure we get a better relative and absolute
tolerance.
The following changes were made
- Renamed iter to iter_start
- Renamed tile_iter to tile_iter_start
- Moved documentation from member variables to getters
- Removed double underscore from extra_iters_before_me variable
- Defined parent header in impl file
- Removed unused inlcudes
There are 2 derived structs based on whether Stream-K is persistent or not.
If it's persistent that means that both the data parallel and Stream-K sections
are data parallel. If it's non-persistent that means that only the
Stream-K section is persistent, while the data parallel section will have
separate workgroups allocated for it. Both structs will have a template
argument for Persistent.
The 2 derived classes will inherit common variables and functions from the
Stream-K TilePartitioner base class. There are additional variables for the
differing data parallel sections that will be added to each derived class,
that are in charge of the indexing/bookkeeping for the data parallel sections.
The only additional function that will differ between the 2 structs is GridSize(),
as the non-persistent will allocate extra workgroups for data parallel.
Unit tests for the derived structs are included.
To better align with the original Stream-K paper, this change implements
a new Stream-K tile partitioner base class. This class will handle the
Stream-K setup that is common to both a persistent and non-persistent DP
section. A later change will implement derived classes to handle the
differences between persistent and non-persistent DP.
This change also includes unit tests for the base tile partitioner.
* [CK_TILE] Correct BlockWarps calculation and fix smoke-test in rmsnorm
* Update rmsnorm host reference
* Update tree reduction of rmsnorm for reference host
* Fix cross warp for m > 1 cases
* Add RMSNorm model selectable option for host reference
* Fix save_unquant cases
* Update reference rmsnorm forward function to use enum for model sensitivity
* Update reference rmsnorm calculation for model sensitivity
* Fix m warp for layernorm
* Adjust parameter of reference for twoPass
* Fix clang format
* Run clang-format-overwrite.sh to fix formating issue
* fix clang format
---------
Co-authored-by: MHYang <mengyang@amd.com>
Co-authored-by: illsilin_amdeng <Illia.Silin@amd.com>
Co-authored-by: ThomasNing <thomas.ning@amd.com>
* Enable the adapted LDS B layout for Row-Major
* fix formatting
* Implement specialized col-major A LDS block descriptor
* Fix formatting
* Use VecLoadSize for AK1/BK1
* Fix some thread access pattern values
* Use GetVectorSizeA for A
* Fix formatting
* Add extra condition to avoid division by zero
* disable layout for wave32
* remove extra else
* fix formatting
* Fix formatting
* Rename one remaining TileDistributionEncodingPattern2D
* Use integer ceil division
* revert remod.py changes
* also revert utility.hpp
* use getA/BTileAccessPattern everywhere
* use integer_divide_ceil for AK0 too
---------
Co-authored-by: Adam Osewski <19374865+aosewski@users.noreply.github.com>
Co-authored-by: Adam Osewski <Adam.Osewski@amd.com>
* Initial commit. create batched_contraction_kernel file
* initial problem definition
* implement initial example to launch kernel
* add universal gemm to contraction. initial phase
* complete implementation for special case all Dims are 1 and no Ds
* clean code
* initial changes to support multi dimensional G
* more progress in implementing multiple G
* tmp commit
* manage dynamic NumDimG in kernel
* improving example for multi M,N,K,G handling. start generalizing kernel. it is a temporary commit
* implement the example for general Multi dimension G M N K and test different reference calculation algorithms
* 2 functions for reference using multi dimensional and flat indexing
* clean the code for muti dimentional G, M, N, K contraction and add some logs
* Add Make descriptor function in kernel for merging Ms, Ns, Ks for A, B, E
* some cleaning on kernel
* clean the code for calculating the offsets from flatten batch number
* Start adding MultiD support to kernel and example
* more changes to manage multi D in kernel and example
* manage passing multi d to kernel and testing.
* complete multi D support in kernel. modify example code to support it
* Correct algorithm to calc the correct offset values for D tensor batches and some code cleaning
* Minor fix
* Generalize example code for variable NumD tensors and apply cleanup based on review feedback
* Refactored code and addressed review feedback
* refactoring, cleaning, add documents, in kernel side and example codes
* Optimize batch offset calculation in kernel
* Inline CalculateBatchOffset in batched contraction kernel, update CHANGELOG.md
---------
Co-authored-by: Adam Osewski <19374865+aosewski@users.noreply.github.com>
* Add initial fp16_mem_128x128x32_2x2x1_32x32x16_NonPersistent test suite
* Account for stride when computing K offsets for A and B tensor
This change ensures that the correct stride is used when computing the K
offsets into the A and B tensors in the Stream-K Kernel's operator()
function. This ensures that the kernel executes correct regardless of
whether A and B are row or column major.
* Move helper code to test_gemm_streamk_util.hpp
* Separate tests into smoke/regression/extended. Add bf16 datatype
* Run clang-format
* Refactor combinatorial macro expansion and naming
* Adjust the initialization values to account for better tolerance on bf16
* Correct BF16 datatypes in comments
* Move the extended tests under the REGRESSION_TESTS label
* Apply suggestions from code review
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
---------
Co-authored-by: Emily Martins <emily.martins@amd.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* debugging
* debugging for prefill shapes
* comment unused code
* fix for prefill shapes
* clearing up the code
* add int4 to universal gemm example
* clang formatted
* adding test for prefill shapes in block scale gemm
* lil improv on the block pipeline
* Address Review Comment
---------
Co-authored-by: ThomasNing <thomas.ning@amd.com>
* reuse local prefetch logic from compute v4 pipeline
add single-tile test
explicit lambda capture
reuse lds block descriptors from base policy for the transposed case
match the test case kernel configuration with compute v4
* add comments
* Pooling 2D/3D with refernce
* Tests & cleanup
- added test for ppoling
- cleanup
- removed 2d example
* Comment resolution
- README added
- example target name rectified
- appropriate arg description and comments added
* clang-format
* appropriate blocksize calc
* modifications for future indexing addition
- instead of transforming views we now transform the descriptors, so
that the same descriptor can be re-used for index tensor in the future
* some basic fixes
* comment resolutions
* comment resolutions
---------
Co-authored-by: Illia Silin <98187287+illsilin@users.noreply.github.com>
* WIP: add memory pipeline boiler plate code that compiles and works for one block
* WIP: tail handling works for memory pipeline
* WIP: numerical errors appears to have gone by adding block_sync_lds()
* fix: numerical error with memory pipeline by adding block_sync_lds() and new tail handler
* refactror: remove debug print statements and lints
* fix: remove redundant sync barriars
* chore: remove lint
* fix: remove unused code from tile handler and remove redundant block_sync_lds()
* fix: correct parent struct name for memory pipeline
* fix: remove static assert check from parent struct and add it to child struct because not all child structs needs to static assert
* fix: defer block sync lds to just before prefill
* [CK TILE GEMM] Support Aquant GEMM with transposeC and preshuffle
When TransposeC and QuantPreshuffle are both true, Aquant generates
correct result.
* [CK TILE GEMM] Support Aquant GEMM with transposeC and preshuffle
- Add unit tests
* Fix bug in is_quantpreshuffle_enabled
* clang format
---------
Co-authored-by: ThomasNing <thomas.ning@amd.com>
* check in pipeline and policy
for async load in mi350, need to make sure TileAccessPattern is warp_raked or block_raked
solve merge conflicts
* fix cmakelists
* make it build
* fix? buffer async fence
* relax fences; it appears it only is needed between pairs of ping-pongs
* remove fences
* remove fences
* cleanup and reformat
* add steps annotations
* comment all pipeline steps / remove unexplainable syncs
* clang-format
* add comment
* cleanup kernel types for test
* fix comment
* fix hardcoded warp size
* faithfully copy block gemm from compute v4 policy to async policy
* make async test gfx950 only
* fix cmake logic
* set separate compile options for async
* refine comment in policy
* try update hotloop scheduler
* cleanup comments
* test more K block sizes
* unhardcode Ks, sort of
* add large odd test case
* fix build for quant
* add comment to hot loop scheduler and rename enum
* reformat
* reword the pipeline description
* reformat
* address review / add static asserts / typo fix
* update changelog
* Use __builtin_amdgcn_readfirstlane for buffer resource in fused_moe
* also do the same for amd_buffer_addressing_builtins.hpp
* merge with develop
* fix clang format
---------
Co-authored-by: Adam Osewski <19374865+aosewski@users.noreply.github.com>
Co-authored-by: ThomasNing <thomas.ning@amd.com>
Co-authored-by: illsilin_amdeng <Illia.Silin@amd.com>
* feat(grouped_gemm_multi_d): add new example that integrates grouped_gemm and multi_d_gemm feature
* refactor: grouped_gemm_multi_d relies on grouped_gemm_kernel
* tests(grouped_gemm): grouped_gemm test suite passes with minor adjustments
* fix: segfault fix by passing correct parameters for d tensors
* style: clang format
* WIP: host code for grouped_gemm_multi_d persistent kernel compiles but segfaults
* feat(grouped_gemm_multi_d): add functionality to run persistant kernel
* feat(grouped_gemm_multi_d): add new example that integrates grouped_gemm and multi_d_gemm feature
* refactor: grouped_gemm_multi_d relies on grouped_gemm_kernel
* tests(grouped_gemm): grouped_gemm test suite passes with minor adjustments
* fix: segfault fix by passing correct parameters for d tensors
* style: clang format
* fix: incorrect validation method and Dtensor layout in test suite
* docs: improved README text based on review comments
* fix: parameterize NumDTensor in GroupedGemmHostArgs and remove lint
* initial commit
* remove extra files
* fixing errors
* updated ReadMe file for mapping of diff quants with diff configs
* addressing review comments
* addressing review comments
* Resolved merge conflicts
* [CK TILE GEMM] Replace get_preshuffle_or with is_quantpreshuffle_enabled
The get_preshuffle_or was not working as expected, which led to incorrect behavior
in the quantization preshuffle process. This change replaces it with the more reliable
is_quantpreshuffle_enabled function to properly determine when preshuffle should be applied.
* initial commit
* debugging
* working fp8 for init constant
* fp8 working with all inits
* updated block level code with comments
* changing the loop iter
* debugging
* debugging
* debugging
* code fix
* code clean up
* clang formatted
* Add comment
* code cleanup
* clang formatted
* merge conflicts fixes
* applying the latest int4 changes to the piepline
* fixing test code for updated traits
* Adding gtest
* review comments addressed
* addressing review comments
* remove c++20 code
* added flush cache changes
---------
Co-authored-by: Cong Ma <congma13@amd.com>
Co-authored-by: root <root@banff-cyxtera-s73-2.ctr.dcgpu>
* Change the return type of run_gemm_combinations in the basic tests
* Change the return type of run_gemm_combinations in the universal tests
* Add universal GEMM tests for bf16 x pk_i4 and fp16 x pk_i4
* Add universal GEMM test for fp8 x pk_i4
* Add basic GEMM tests for bf16 x pk_i4, fp16 x pk_i4 and fp8 x pk_i4.
* Add missing GemmTypeConfig<ck_tile::fp8_t, ck_tile::pk_int4_t, ck_tile::half_t>
* Add missing GemmTypeConfig<ck_tile::bf16_t, ck_tile::pk_int4_t, ck_tile::bf16_t>
* No need for utility in test_ck_tile_elementwise_1d
* Fix conversion from pk_int4x4_t to bf16x8_t in PassThroughPack8
* Avoid union-based type punning in float_to_bf16_truc_raw to make it constexpr compliant
* For consistency also make float_to_bf16_truc_nan_raw constexpr compliant by removing the union
* Use a static_cast to bfloat16_t only when CK_TILE_USE_LLVM_BUILTIN_BF16 is enforced
* Convert from float to bf16 during compilation rather than using magic values
* Fix conversion from pk_int4x4_t to fp8x8_t in PassThroughPack8
* Comment out the basic test for fp16 x pk_i4 as it does not pass
* Add missing GemmTypeConfig<ck_tile::bf8_t, ck_tile::pk_int4_t, ck_tile::half_t>
* Fix conversion from pk_int4x4_t to bf8x8_t in PassThroughPack8
* Add basic and universal GEMM tests for bf8 x pk_i4
* Switch back to amd_assembly_i4_to_fp8x8 in PassThroughPack8 as it works now
* Switch back to amd_assembly_i4_to_bf8x8 in PassThroughPack8 as it works now
* Remove the inefficient fallbacks for fp8 and bf8 in elementwise/unary_element_wise_operation.hpp
* Use explicit macros for enabling and disabling the the constexpr lookup based converters
* Fix two failing tests
* Avoid union-based type punning in float_to_bf16_rtn_raw to make it constexpr compliant
* Use float_to_bf16_rtn_raw instead of float_to_bf16 to create the bf16 lookup table for use in conversions from pk_int4 to bf16
* On ROCm 7.0.1 we need an explicit cast to from uint16_t to bf16_t
* Support 16x16 (MFMA, WMMA) and 32x32 (MFMA) tiles in fwd and bwd BlockDropout
Add comments with dropout implementation details
Fix performance regression of fwd+dropout
* Remove some usage of type punning (reinterpret_cast with ref or ptr) in Philox;
* "scalarize" seed and offset, they may come either from kernel args or from device memory
(presumably loaded with vector loads).
These changes help the compiler to procude more optimal code and reduce register spilling.
Use WarpGemmDispatcher instead of explicit WarpGemmMfma... to get CWarpDstrEncoding
Use code based on BlockDropout in BlockDropoutBwd
Refactor BlockDropout (fwd)
Implement BlockDropout (fwd) for WMMA
Originally BlockDropout only supported 32x32 tiles (IsWG32 = true),
this version supports 16x16 tiles.
If MPerBlock > MWarp * 16, it can generate numbers for two 16x16 tiles, similarly
to BlockDropoutBwd.
Implement BlockDropoutBwd for WMMA
Remove MakeRandValLds* functions unused in BlockDropoutBwd
Remove unused Run overload from BlockDropoutBwd
* Fix regression with philox seed and offset when they exceed 32-bit int
__builtin_amdgcn_readfirstlane works with 32-bit values, seed and offset
are 64-bit so they get truncated.
* Add F32 MFMA warp gemms
* Support f32 in fwd FMHA
* Implement transpose_vectors for 4-byte types (float)
* Fix unexpected implicit f32->uint32 cast in buffer_store<4>
__builtin_amdgcn_raw_buffer_store_b32 expects unsigned int but float was passed (implicitly casted to uint).
mbuf_t types in other buffer_store<> are changed for consistency.
* Support F32 in bwd FMHA
hdim = 256 is disabled for now because it uses too much memory on gfx90a
* Support Headdim = 48 (divisible by 16) in fwd
* Add fp32-specific receipts (800 and 801)
* Tune fwd tiles
* Tune bwd tiles
* Use small tiles only for small seqlen_q
* Fix after rebasing
* Fix selection of a fallback tile based on bm0
The assumption that the largest bm0 == 128 is not always true for
current fp32 tiles.
* Remove constraints and adjust filtering for fp32
Custom constraints are no longer needed because now the smallest tile
is selected automtically based on seqlen_q.
Filters related to qr_async_trload disabled valid fp32 tiles.
* Add fp32 tests
* Make splitkv and appendkv compile for fp32 only
There are no instances yet, but API still must compile when only fp32 is
requested.
* Remove unimportant f32 instances
* Add test_ck_tile_fmha_*_fp32 to REGRESSION_TESTS
* Replace magic numbers with a constant, improve comments for dropout
* Update changelog
* Fix condition that dq_acc must be set to zero when mask is used
The change was introduced in #2799
* Replace warp_uniform with recently added amd_wave_read_first_lane
* Add hdim = 96 and 192 to fwd
* Fix validation of rotary embedding with time_kernel_
When rotary embedding is used, the appendkv kernel modifies the q tensor
(multiple times when time_kernel_ is set). We need to reset the q buffer
and rerun all kernels.
* Fix synchronization issue in splitkv combine pipeline
Different warps can read and then rewrite the same values of lse_acc_lds.
Sometimes warps progress at different speeds, one warp can rewrite
values that are still being read by another warp.
Running the tests multiple times and, preferably, with multiple
processes on the same GPU helps to trigger this issue:
bin/test_ck_tile_fmha_fwd_fp16 --gtest_repeat=-1 --gtest_shuffle --gtest_throw_on_failure --gtest_filter="TestCkTileFmhaFwd/*KV*"
* feat(grouped_gemm_multi_d): add new example that integrates grouped_gemm and multi_d_gemm feature
* feat: generalized grouped_gemm_kernel.hpp
* feat: generalized grouped_gemm_kernel.hpp even further by removing hardcoded 0
* refactor: grouped_gemm_multi_d relies on grouped_gemm_kernel
* tests(grouped_gemm): grouped_gemm test suite passes with minor adjustments
* fix: segfault fix by passing correct parameters for d tensors
* docs: add multi d info and trim down outdated content
* tests: add unit tests for grouped_gemm_multi_d and minor changes in grouped_gemm related test for compatibility
* style: clang format
* fix: incorrect validation method and Dtensor layout in test suite
* * [CK_TILE] Add sequence padding and variable length support in fmha (and v3)
- Group Mode Padding: Introduces the `-s_qpad` argument to support
physically padded layouts. Kernels now use padded start pointers
(`seqstart_padded_*_ptr`) for memory addressing.
- Batch Mode Variable Length: Adds `-q_eff_lens` and `-kv_eff_lens`
arguments for efficient processing of variable-length sequences by
passing cumulative effective lengths (`cu_seqlen_*_ptr`) to the kernel.
- FMHA examples: Support padding and variable length both in
group and batch mode. Dispatcher is updated as well (dispatch to
kPadSeqLenK enabled pipeline).
- New padding test cases: Add padding test cases to `smoke_test_fwd.sh` and
`test_fmha_fwd.inc`, and add benchmarks to `benchmark_fwd.sh` and
`benchmark_fwd_v3.sh` as well. These test cases and benchmarks that
specifically validate/benchmark the new padding and variable-length
functionalities in both group and batch modes.
* [CK_TILE] Fix build error in fmha unit tests
* [CK_TILE] add mqa, gqa to sequence padding unit tests
* [CI_TILE] Reduce the number of padding seqlen unit tests in FMHA to avoid timeouts in CI
* [CK_TILE] remove unnecessary MageKArgs overload in FmhaFwdV3Kernel and FmhaFwdKernel
