streamk_gemm_tile_partitioner_impl.hpp Source File#
streamk_gemm_tile_partitioner_impl.hpp
Go to the documentation of this file.
32 }
40 iters_per_sk_cta_ = sk_ctas_ ? total_sk_iters_ / sk_ctas_ : 0;
71}
147}
151StreamKTilePartitionerBase<BlockGemmShapeType, ReductionStrategyType>::get_dp_tiles() const noexcept
152{
Definition tile/core/algorithm/cluster_descriptor.hpp:13
__device__ uint32_t amd_wave_read_first_lane(uint16_t v)
Definition tile/core/arch/amd_buffer_addressing.hpp:35
CK_TILE_HOST_DEVICE constexpr auto integer_divide_ceil(X x, Y y)
Definition tile/core/numeric/math.hpp:149
CK_TILE_HOST_DEVICE constexpr auto make_tuple(Xs &&... xs)
Definition tile/core/container/tuple.hpp:360
index_t sk_start_block_idx_
Definition streamk_gemm_tile_partitioner.hpp:328
index_t dp_ctas_
Definition streamk_gemm_tile_partitioner.hpp:326
StreamKTilePartitioner_v2(ck_tile::index_t m, ck_tile::index_t n, ck_tile::index_t k, ck_tile::index_t grid)
Definition streamk_gemm_tile_partitioner_impl.hpp:290
index_t dp_start_block_idx_
Definition streamk_gemm_tile_partitioner.hpp:327
StreamKTilePartitioner_v2(ck_tile::index_t m, ck_tile::index_t n, ck_tile::index_t k, ck_tile::index_t grid)
Definition streamk_gemm_tile_partitioner_impl.hpp:246
index_t dp_tiles_per_cta_
Definition streamk_gemm_tile_partitioner.hpp:275
index_t extra_dp_tiles_
Definition streamk_gemm_tile_partitioner.hpp:276
Template for the Stream-K tile partitioner derived struct.
Definition streamk_gemm_tile_partitioner.hpp:230
CK_TILE_HOST_DEVICE index_t get_sk_tiles() const noexcept
Returns the number of tiles in the C tensor that will use the Stream-K approach.
Definition streamk_gemm_tile_partitioner_impl.hpp:158
CK_TILE_HOST_DEVICE index_t get_total_dp_iters() const noexcept
Returns the total number of DP iterations.
Definition streamk_gemm_tile_partitioner_impl.hpp:204
CK_TILE_HOST_DEVICE index_t get_flags_buffer_size() const noexcept
Calculates the total space needed for the flags buffer.
Definition streamk_gemm_tile_partitioner_impl.hpp:57
CK_TILE_HOST_DEVICE index_t get_num_tiles() const noexcept
Returns the number of macro tiles in the C tensor.
Definition streamk_gemm_tile_partitioner_impl.hpp:136
CK_TILE_HOST_DEVICE index_t get_partials_buffer_size(index_t acc_element_bytes) const noexcept
Calculates the total space needed for the partials buffer.
Definition streamk_gemm_tile_partitioner_impl.hpp:49
CK_TILE_DEVICE void get_iter_boundaries(index_t &iter_start, index_t &iter_end, index_t cta_idx) const noexcept
Calculates the start and end iteration given the cta_idx.
Definition streamk_gemm_tile_partitioner_impl.hpp:65
CK_TILE_DEVICE void get_tile_boundaries(index_t &tile_iter_start, index_t &tile_iter_end, index_t tile_idx) const noexcept
Calculates the starting and ending tile boundaries for the given 1D tile index.
Definition streamk_gemm_tile_partitioner_impl.hpp:83
CK_TILE_DEVICE auto get_output_tile_index(index_t tile_idx) const noexcept -> tuple< index_t, index_t >
Calculates the workgroups 2D tile index in the C tensor given the 1D tile index.
Definition streamk_gemm_tile_partitioner_impl.hpp:108
index_t grid_
Definition streamk_gemm_tile_partitioner.hpp:195
CK_TILE_HOST_DEVICE index_t get_extra_iters() const noexcept
Returns the remainder resulting from total_sk_iters_ divided by sk_ctas_. When this is non-zero,...
Definition streamk_gemm_tile_partitioner_impl.hpp:196
static constexpr index_t KPerBlock
Definition streamk_gemm_tile_partitioner.hpp:29
static CK_TILE_DEVICE index_t get_local_iter_end(index_t tile_iter_start, index_t iter_end, index_t tile_iter_end) noexcept
Calculates the workgroup's non-inclusive end iteration that is local to a tile.
Definition streamk_gemm_tile_partitioner_impl.hpp:100
CK_TILE_HOST_DEVICE index_t get_grid() const noexcept
Returns the maximum number of active workgroups; this is assumed to be number of CUs * occupancy.
Definition streamk_gemm_tile_partitioner_impl.hpp:144
static constexpr StreamKReductionStrategy ReductionStrategy
Definition streamk_gemm_tile_partitioner.hpp:30
CK_TILE_DEVICE index_t get_tile_index(index_t iter_start) const noexcept
Calculates the 1D tile index in the C tensor for a workgroup.
Definition streamk_gemm_tile_partitioner_impl.hpp:75
CK_TILE_HOST_DEVICE index_t get_total_sk_iters() const noexcept
Returns the total number of Stream-K iterations.
Definition streamk_gemm_tile_partitioner_impl.hpp:172
CK_TILE_HOST_DEVICE index_t get_dp_tiles() const noexcept
Returns the number of tiles in the C tensor that will use the data-parallel (DP) approach.
Definition streamk_gemm_tile_partitioner_impl.hpp:151
CK_TILE_HOST_DEVICE index_t get_iters_per_tile() const noexcept
Returns the total number of iterations per tile in the C tensor. In other words, this is the total nu...
Definition streamk_gemm_tile_partitioner_impl.hpp:180
CK_TILE_HOST_DEVICE index_t get_n() const noexcept
Returns the n dimension for the GEMM problem.
Definition streamk_gemm_tile_partitioner_impl.hpp:212
static constexpr index_t NPerBlock
Definition streamk_gemm_tile_partitioner.hpp:28
static constexpr index_t MPerBlock
Definition streamk_gemm_tile_partitioner.hpp:27
index_t num_tiles_
Definition streamk_gemm_tile_partitioner.hpp:194
CK_TILE_HOST_DEVICE index_t get_workspace_size(index_t acc_element_bytes) const noexcept
Calculates the total space needed for the partials and flags buffers.
Definition streamk_gemm_tile_partitioner_impl.hpp:120
StreamKTilePartitionerBase(index_t m, index_t n, index_t k, index_t grid)
Definition streamk_gemm_tile_partitioner_impl.hpp:8
CK_TILE_HOST_DEVICE index_t get_iters_per_sk_cta() const noexcept
Returns the total number of Stream-K iterations for each sk_cta. This is the lower bound (i....
Definition streamk_gemm_tile_partitioner_impl.hpp:188
CK_TILE_HOST_DEVICE index_t get_sk_ctas() const noexcept
Returns the number of workgroups that will participate in Stream-K in the sk_tiles_.
Definition streamk_gemm_tile_partitioner_impl.hpp:165
static CK_TILE_DEVICE index_t get_local_iter(index_t iter_start, index_t tile_iter_start) noexcept
Calculates the workgroup's starting iteration that is local to a tile.
Definition streamk_gemm_tile_partitioner_impl.hpp:92
CK_TILE_HOST index_t estimate_num_wgs_per_tile() const noexcept
Returns an estimate of the number of workgroups writing to the same macro tile in C.
Definition streamk_gemm_tile_partitioner_impl.hpp:219
index_t dp_tiles_
Definition streamk_gemm_tile_partitioner.hpp:196
Definition tile/core/container/tuple.hpp:192