MCPcopy Create free account

hub / github.com/Tencent/hpc-ops / functions

Functions10,542 in github.com/Tencent/hpc-ops

⨍ Functions 10,542◇ Types & classes 8,025↳ Endpoints 13
↓ 3 callersMethodbegin_row
3rd/cutlass/include/cutlass/epilogue/threadblock/fusion/visitor_store.hpp:315
↓ 3 callersFunctionblock
3rd/cutlass/include/cute/util/debug.hpp:121
↓ 3 callersMethodclear_mask
Clears the predicate set efficiently
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:924
↓ 3 callersMethodconvert_from_float
3rd/cutlass/include/cutlass/float8.h:1073
↓ 3 callersFunctioncooperative_gemm_predication
3rd/cutlass/include/cute/algorithm/cooperative_gemm.hpp:166
↓ 3 callersFunctioncopy_unpack
3rd/cutlass/include/cute/atom/copy_traits.hpp:111
↓ 3 callersFunctioncrd2crd
3rd/cutlass/include/cute/stride.hpp:254
↓ 3 callersMethoddiv
Computes integer division using precomputed values. This is computationally inexpensive.
3rd/cutlass/include/cutlass/fast_math.h:388
↓ 3 callersMethodell_add_mask
add mask for small tiles in ELL
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:940
↓ 3 callersMethodenable
< CUTLASS_HOST_DEVICE enables all accesses guarded by mask
3rd/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h:179
↓ 3 callersMethodenable_mask
Clears the predicate set efficiently
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:928
↓ 3 callersMethodend
3rd/cutlass/include/cute/container/array.hpp:129
↓ 3 callersMethodend_row
3rd/cutlass/include/cutlass/epilogue/threadblock/fusion/visitor_store.hpp:341
↓ 3 callersMethodend_step
Called after all accumulator elements have been visited
3rd/cutlass/include/cutlass/epilogue/threadblock/fusion/visitor_2x.hpp:111
↓ 3 callersFunctionfill_workspace
3rd/cutlass/include/cutlass/workspace.h:97
↓ 3 callersFunctionfind
3rd/cutlass/include/cute/container/tuple.hpp:254
↓ 3 callersFunctionflatten
3rd/cutlass/include/cute/tensor_impl.hpp:582
↓ 3 callersFunctionget
3rd/cutlass/include/cute/layout.hpp:497
↓ 3 callersMethodget
Returns a pointer
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:883
↓ 3 callersMethodget
Returns a pointer
3rd/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op.h:299
↓ 3 callersFunctiongetParams
3rd/cutlass/include/cutlass/conv/threadblock/conv2d_fprop_activation_tile_access_iterator_analytic.h:177
↓ 3 callersMethodget_batch_idx
Obtains the batch index
3rd/cutlass/include/cutlass/gemm/threadblock/threadblock_swizzle_streamk.h:653
↓ 3 callersFunctionget_grid_shape
3rd/cutlass/include/cutlass/transform/kernel/sm90_sparse_gemm_compressor.hpp:183
↓ 3 callersMethodget_grid_shape
3rd/cutlass/include/cutlass/gemm/kernel/sm100_tile_scheduler.hpp:204
↓ 3 callersMethodget_layoutA_TV
3rd/cutlass/include/cute/atom/mma_atom.hpp:414
↓ 3 callersMethodget_layoutB_TV
3rd/cutlass/include/cute/atom/mma_atom.hpp:436
↓ 3 callersMethodget_mask
Gets the mask
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:936
↓ 3 callersMethodget_mask
< Sets the mask
3rd/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h:1060
↓ 3 callersFunctionget_max_cta_occupancy
3rd/cutlass/include/cutlass/gemm/kernel/tile_scheduler_params.h:54
↓ 3 callersFunctionget_sm_major_version
src/utils/utils.cc:29
↓ 3 callersMethodget_stride
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:893
↓ 3 callersMethodis_last_tile
3rd/cutlass/include/cutlass/gemm/kernel/static_tile_scheduler.hpp:196
↓ 3 callersMethodis_producer_load_needed
3rd/cutlass/include/cutlass/epilogue/collective/sm90_epilogue_array_tma_warpspecialized.hpp:473
↓ 3 callersMethodis_producer_load_needed
3rd/cutlass/include/cutlass/epilogue/collective/sm100_epilogue_tma_warpspecialized.hpp:447
↓ 3 callersMethodis_producer_load_needed
3rd/cutlass/include/cutlass/epilogue/collective/sm100_epilogue_array_tma_warpspecialized.hpp:464
↓ 3 callersMethodload
(self, input_rank, residual, weight)
benchmark/fuse_allreduce_rmsorm/benchmark_fuse_allreduce_rmsnorm.py:299
↓ 3 callersMethodload
Load
3rd/cutlass/include/cutlass/block_striped.h:157
↓ 3 callersFunctionload_with_byte_offset
Loads a fragment from memory
3rd/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h:310
↓ 3 callersMethodload_with_byte_offset
Loads a fragment from memory
3rd/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator.h:1465
↓ 3 callersFunctionload_with_pointer_offset
Loads a fragment from memory with additional logical offset
3rd/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op_tile_iterator_sm80.h:284
↓ 3 callersFunctionload_with_pointer_offset
Load
3rd/cutlass/include/cutlass/epilogue/warp/tile_iterator_tensor_op.h:210
↓ 3 callersMethodload_with_pointer_offset
Loads a fragment from memory
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_iterator.h:871
↓ 3 callersFunctionmake_swizzle_ptr
3rd/cutlass/include/cute/pointer_swizzle.hpp:125
↓ 3 callersFunctionmake_tma_atom_im2col
3rd/cutlass/include/cute/atom/copy_traits_sm90_im2col.hpp:574
↓ 3 callersMethodmantissa
Returns the mantissa
3rd/cutlass/include/cutlass/tfloat32.h:185
↓ 3 callersMethodmantissa
Returns the mantissa
3rd/cutlass/include/cutlass/half.h:487
↓ 3 callersMethodmantissa
Returns the mantissa
3rd/cutlass/include/cutlass/bfloat16.h:223
↓ 3 callersMethodmax_dim_index
Returns the index of the dimension with greatest value
3rd/cutlass/include/cutlass/coord.h:141
↓ 3 callersMethodmin
Least positive value
3rd/cutlass/include/cutlass/float8.h:1352
↓ 3 callersMethodoperator++
3rd/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_tensor_op_sm70.h:498
↓ 3 callersMethodoperator++
3rd/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op_sm80.h:588
↓ 3 callersMethodpipeline
3rd/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler.hpp:62
↓ 3 callersFunctionpopcount
3rd/cutlass/include/cutlass/functional.h:70
↓ 3 callersFunctionproducer_commit
NOP for TMA based mainloop
3rd/cutlass/include/cutlass/pipeline/sm90_pipeline.hpp:561
↓ 3 callersMethodproducer_commit
3rd/cutlass/include/cutlass/pipeline/sm100_pipeline.hpp:1156
↓ 3 callersFunctionproducer_try_acquire
3rd/cutlass/include/cutlass/pipeline/sm100_pipeline.hpp:119
↓ 3 callersMethodproduct
Returns the product of all elements
3rd/cutlass/include/cutlass/coord.h:319
↓ 3 callersMethodpromote_residue_if_needed
3rd/cutlass/include/cutlass/gemm/collective/fp8_accumulation.hpp:182
↓ 3 callersFunctionprovider_key
(provider)
benchmark/route_gemm/benchmark_gemm_bf16xfp32.py:273
↓ 3 callersMethodreset
Updates only the pointer
3rd/cutlass/include/cutlass/tensor_ref.h:249
↓ 3 callersFunctionsafe_div
3rd/cutlass/include/cute/swizzle.hpp:365
↓ 3 callersFunctionsafe_div
3rd/cutlass/include/cute/numeric/math.hpp:307
↓ 3 callersFunctionscaled_fp8_quant
Quantize a tensor to FP8 with one per-tensor scale.
hpc/act.py:108
↓ 3 callersFunctionset_block_rank
Set the destination block-ID in cluster for a given SMEM Address
3rd/cutlass/include/cute/arch/cluster_sm90.hpp:166
↓ 3 callersMethodset_iteration_index
Overrides the internal iteration index
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:866
↓ 3 callersMethodset_iteration_index
Overrides the internal iteration index
3rd/cutlass/include/cutlass/transform/threadblock/regular_tile_access_iterator_tensor_op.h:289
↓ 3 callersMethodset_kgroup_index
Notify the iterator which k-group it is currently pointing to. This does not advance the iterator. Rather, it overrides its internal tracking with co
3rd/cutlass/include/cutlass/gemm/warp/mma_complex_tensor_op_tile_iterator_sm80.h:568
↓ 3 callersMethodset_kgroup_index
Notify the iterator which k-group it is currently pointing to. This does not advance the iterator. Rather, it overrides its internal tracking with co
3rd/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_sm70.h:1115
↓ 3 callersMethodset_kgroup_index
Notify the iterator which k-group it is currently pointing to. This does not advance the iterator. Rather, it overrides its internal tracking with co
3rd/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator.h:2118
↓ 3 callersMethodset_mask
Sets the predicate mask, overriding value stored in predicate iterator
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:932
↓ 3 callersFunctionshiftr
3rd/cutlass/include/cute/swizzle.hpp:349
↓ 3 callersFunctionsilu
(x)
tests/test_act.py:22
↓ 3 callersFunctionslice_accumulator
3rd/cutlass/include/cutlass/gemm/collective/sm100_mma_array_warpspecialized_blockwise_scaling.hpp:557
↓ 3 callersFunctionslice_accumulator
3rd/cutlass/include/cutlass/gemm/collective/sm100_mma_array_warpspecialized.hpp:470
↓ 3 callersFunctionslice_and_offset
3rd/cutlass/include/cute/layout.hpp:698
↓ 3 callersMethodstorage
3rd/cutlass/include/cutlass/transform/kernel/sm90_sparse_gemm_compressor.hpp:247
↓ 3 callersMethodstore
Store
3rd/cutlass/include/cutlass/block_striped.h:186
↓ 3 callersFunctionstore_with_byte_offset
Stores a fragment to memory
3rd/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h:385
↓ 3 callersMethodstore_with_byte_offset
Store a fragment to memory
3rd/cutlass/include/cutlass/transform/threadblock/predicated_tile_iterator.h:1483
↓ 3 callersFunctionstore_with_pointer_offset
Store
3rd/cutlass/include/cutlass/epilogue/warp/tile_iterator_tensor_op.h:192
↓ 3 callersMethodstore_with_pointer_offset
Store a fragment to memory
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_iterator.h:899
↓ 3 callersMethodstride
Returns the layout object's stride vector
3rd/cutlass/include/cutlass/tensor_ref.h:291
↓ 3 callersFunctionsynclog_print
3rd/cutlass/include/cutlass/arch/synclog.hpp:958
↓ 3 callersFunctionsynclog_print_wgmma_desc
3rd/cutlass/include/cutlass/arch/synclog.hpp:268
↓ 3 callersFunctiontensormaps_init
3rd/cutlass/include/cutlass/gemm/collective/sm100_mma_array_warpspecialized.hpp:756
↓ 3 callersMethodthrfrg_C
3rd/cutlass/include/cute/atom/mma_atom.hpp:250
↓ 3 callersMethodthrottle_pipeline
3rd/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler.hpp:63
↓ 3 callersMethodtransform_init
3rd/cutlass/include/cutlass/gemm/collective/sm100_mma_warpspecialized_mixed_input.hpp:1062
↓ 3 callersFunctionudiv128
3rd/cutlass/include/cutlass/uint128.h:95
↓ 3 callersFunctionumma_arrive
3rd/cutlass/include/cutlass/arch/barrier.h:796
↓ 3 callersFunctionumma_arrive_multicast_2x1SM
UMMA arrive for MMA_2x1SM + TMA_LOAD_MULTICAST combination
3rd/cutlass/include/cutlass/arch/barrier.h:847
↓ 3 callersMethodvalid
Returns whether access is valid or not
3rd/cutlass/include/cutlass/transform/threadblock/ell_predicated_tile_access_iterator.h:946
↓ 2 callersMethodBroadcastFd
src/communicator/communicator.cc:110
↓ 2 callersMethodGetDeviceId
src/communicator/multicast_communicator.cc:48
↓ 2 callersFunctionLaneId
Computes laneId within a warp
3rd/cutlass/include/cutlass/arch/arch.h:50
↓ 2 callersFunctionLinearCombinationClampIsHeavy
Single source of truth for whether to unroll for `LinearCombinationClamp()`
3rd/cutlass/include/cutlass/epilogue/thread/linear_combination_clamp.h:56
↓ 2 callersFunctionLinearCombinationRelu0IsHeavy
Single source of truth for whether to unroll for `LinearCombinationClamp()`
3rd/cutlass/include/cutlass/epilogue/thread/linear_combination_relu0.h:60
↓ 2 callersFunctionLinearCombinationReluIsHeavy
Single source of truth for whether to unroll for `LinearCombinationClamp()`
3rd/cutlass/include/cutlass/epilogue/thread/linear_combination_relu.h:57
↓ 2 callersFunctionOutputTileThreadMapDesc
3rd/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator_params.h:126
↓ 2 callersFunctionPredicatedScaleBiasVectorAccessIterator
Construct a PredicatedTileAccessIterator with zero threadblock offset
3rd/cutlass/include/cutlass/conv/threadblock/predicated_scale_bias_vector_access_iterator.h:194
← previousnext →701–800 of 10,542, ranked by callers