MCPcopy Create free account

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

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

↓ 74 callersFunctionfor_each
3rd/cutlass/include/cute/algorithm/tuple_algorithms.hpp:172
↓ 74 callersMethodlayout_b
3rd/cutlass/include/cute/layout_composed.hpp:83
↓ 74 callersFunctionshape_div
3rd/cutlass/include/cute/int_tuple.hpp:384
↓ 73 callersMethodempty
3rd/cutlass/include/cute/container/array.hpp:153
↓ 73 callersMethodretile_D
3rd/cutlass/include/cute/atom/copy_atom.hpp:395
↓ 72 callersMethodadd_pointer_offset
Adds a pointer offset in units of Element
3rd/cutlass/include/cutlass/conv/threadblock/conv2d_tile_iterator.h:114
↓ 72 callersFunctiontma_partition
3rd/cutlass/include/cute/atom/copy_traits_sm90_tma.hpp:1401
↓ 71 callersMethodget_host_problem_shape
3rd/cutlass/include/cutlass/gemm/group_array_problem_shape.hpp:116
↓ 71 callersMethodload
Load addend source fragment from global memory
3rd/cutlass/include/cutlass/epilogue/threadblock/epilogue.h:271
↓ 71 callersMethodproducer_commit
3rd/cutlass/include/cutlass/pipeline/sm90_pipeline.hpp:1114
↓ 70 callersFunctiontile_to_mma_shape
3rd/cutlass/include/cute/atom/mma_traits_sm100.hpp:114
↓ 70 callersMethodtranspose
3rd/cutlass/include/cutlass/matrix.h:138
↓ 69 callersFunctionswap
3rd/cutlass/include/cutlass/platform/platform.h:838
↓ 69 callersMethodvalid
Returns whether access is valid or not
3rd/cutlass/include/cutlass/transform/threadblock/predicated_tile_access_iterator.h:295
↓ 68 callersMethodat
Returns a reference to the element at a given Coord
3rd/cutlass/include/cutlass/thread/matrix.h:164
↓ 68 callersFunctionfma
3rd/cutlass/include/cute/arch/simd_sm100.hpp:77
↓ 68 callersMethodwait
3rd/cutlass/include/cutlass/arch/barrier.h:370
↓ 67 callersFunctionflat_divide
3rd/cutlass/include/cute/layout.hpp:1633
↓ 66 callersMethodwork_tile_to_cta_coord
3rd/cutlass/include/cutlass/gemm/kernel/sm100_tile_scheduler.hpp:372
↓ 64 callersMethodproducer_get_barrier
3rd/cutlass/include/cutlass/pipeline/sm90_pipeline.hpp:1136
↓ 63 callersFunctionwait_on_dependent_grids
Issuing the griddepcontrol.wait instruction enforces no global memory access prior to this istruction. This ensures the correctness of global memory a
3rd/cutlass/include/cutlass/arch/grid_dependency_control.h:94
↓ 59 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/tile_iterator_planar_complex.h:238
↓ 58 callersFunctionget_grid_shape
3rd/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler_stream_k.hpp:360
↓ 57 callersMethodadvance
3rd/cutlass/include/cutlass/gemm/kernel/grouped_problem_visitor.h:149
↓ 57 callersMethodget
3rd/cutlass/include/cutlass/array_subbyte.h:148
↓ 56 callersMethodpartition_A
3rd/cutlass/include/cute/atom/mma_atom.hpp:476
↓ 53 callersFunctionfilter
3rd/cutlass/include/cute/layout.hpp:934
↓ 53 callersFunctionget_work_k_tile_count
3rd/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler_stream_k.hpp:692
↓ 53 callersFunctioninitialize_workspace
3rd/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler.hpp:145
↓ 53 callersMethodreset
Replaces the managed object, deleting the old object.
3rd/cutlass/include/cutlass/platform/platform.h:806
↓ 52 callersFunctionclear
3rd/cutlass/include/cute/container/array.hpp:352
↓ 52 callersFunctioncp_async_fence
Establishes an ordering w.r.t previously issued cp.async instructions. Does not block.
3rd/cutlass/include/cutlass/arch/memory_sm80.h:440
↓ 52 callersFunctionfloat
Converts to float
3rd/cutlass/include/cutlass/exmy_base.h:1040
↓ 51 callersFunctionconst_min
3rd/cutlass/include/cutlass/fast_math.h:752
↓ 49 callersFunctionmake_zip_tensor
3rd/cutlass/include/cute/tensor_zip.hpp:157
↓ 49 callersFunctionstride
Returns stride as MatrixCoord
3rd/cutlass/include/cutlass/conv/conv3d_problem_size.h:319
↓ 49 callersMethodsynchronize
3rd/cutlass/include/cutlass/transform/collective/sm90_wgmma_transpose.hpp:124
↓ 49 callersFunctiontma_descriptor_cp_fence_release
3rd/cutlass/include/cute/arch/copy_sm90_desc.hpp:423
↓ 49 callersFunctiontma_descriptor_replace_addr_in_shared_mem
Replace tensor pointer by bringing the tensormap from GMEM into the shared memory
3rd/cutlass/include/cute/arch/copy_sm90_desc.hpp:341
↓ 48 callersMethodat
Returns a reference to the element at a given Coord
3rd/cutlass/include/cutlass/tensor_ref.h:321
↓ 47 callersMethodget_problem_shape
3rd/cutlass/include/cutlass/gemm/group_array_problem_shape.hpp:101
↓ 47 callersMethodpartition_B
3rd/cutlass/include/cute/atom/mma_atom.hpp:487
↓ 46 callersMethodlayout
3rd/cutlass/include/cute/tensor_impl.hpp:192
↓ 46 callersMethodlayout_a
3rd/cutlass/include/cute/layout_composed.hpp:71
↓ 46 callersMethodstore
Stores a fragment to memory
3rd/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h:988
↓ 46 callersFunctiontma_descriptor_fence_acquire
3rd/cutlass/include/cute/arch/copy_sm90_desc.hpp:457
↓ 45 callersMethodtensor
3rd/cutlass/include/cute/tensor_impl.hpp:162
↓ 44 callersMethodsize
Static method that returns the packed logical size of any TensorExtent
3rd/cutlass/include/cutlass/conv/convnd_problem_shape.hpp:490
↓ 43 callersMethoddata
3rd/cutlass/include/cute/tensor_impl.hpp:180
↓ 42 callersFunctionget_tiled_cta_shape_mnl
3rd/cutlass/include/cutlass/gemm/kernel/sm90_tile_scheduler_stream_k.hpp:352
↓ 42 callersMethodlayout
3rd/cutlass/include/cute/layout.hpp:119
↓ 42 callersFunctionsync
3rd/cutlass/include/cutlass/arch/barrier.h:233
↓ 42 callersFunctionwarpgroup_arrive
3rd/cutlass/include/cute/arch/mma_sm90_gmma.hpp:47
↓ 42 callersFunctionwarpgroup_commit_batch
Marks the commit point for one or more sized batch of warpgroup MMAs.
3rd/cutlass/include/cute/arch/mma_sm90_gmma.hpp:74
↓ 41 callersMethodadd_tile_offset
Advances an iterator along logical dimensions of matrix in units of whole threadblock tiles
3rd/cutlass/include/cutlass/conv/threadblock/predicated_scale_bias_vector_iterator.h:347
↓ 41 callersFunctionright_inverse
3rd/cutlass/include/cute/layout.hpp:1263
↓ 41 callersFunctionto_gemm_coord
3rd/cutlass/include/cutlass/gemm_coord.hpp:47
↓ 41 callersMethodview
Returns a TensorRef pointing to the first element of the tensor.
3rd/cutlass/include/cutlass/thread/matrix.h:152
↓ 41 callersMethodx
Accesses the x element of the imaginary part of the quaternion
3rd/cutlass/include/cutlass/quaternion.h:152
↓ 41 callersMethody
Accesses the y element of the imaginary part of the quaternion
3rd/cutlass/include/cutlass/quaternion.h:164
↓ 41 callersMethodz
Accesses the z element of the imaginary part of the quaternion
3rd/cutlass/include/cutlass/quaternion.h:176
↓ 40 callersMethodget
Returns a pointer to the vector starting at the current coordinate
3rd/cutlass/include/cutlass/conv/threadblock/conv3d_fprop_filter_tile_access_iterator_optimized.h:233
↓ 40 callersMethodget_flat_coord
3rd/cutlass/include/cute/layout.hpp:261
↓ 40 callersMethodget_tiled_shape
Returns the GEMM volume in thread block tiles
3rd/cutlass/include/cutlass/gemm/kernel/params_universal_base.h:201
↓ 40 callersMethodload_tail
Perform a Producer Epilogue to prevent early exit of ctas in a Cluster
3rd/cutlass/include/cutlass/gemm/collective/sm100_mma_warpspecialized.hpp:629
↓ 40 callersMethodoffset
3rd/cutlass/include/cute/layout_composed.hpp:77
↓ 40 callersMethodto_half
Bitcasts to CUDA's half type
3rd/cutlass/include/cutlass/half.h:445
↓ 39 callersMethodend
Exit of the subtile load loop.
3rd/cutlass/include/cutlass/epilogue/fusion/sm90_visitor_tma_warpspecialized.hpp:154
↓ 39 callersMethodload
Loads a fragment from memory
3rd/cutlass/include/cutlass/conv/threadblock/conv2d_tile_iterator.h:166
↓ 37 callersFunctionload_with_pointer_offset
Loads a fragment from memory
3rd/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_tensor_op_sm70.h:201
↓ 37 callersFunctionlogical_divide
3rd/cutlass/include/cute/layout.hpp:1557
↓ 37 callersFunctionround_up
Returns the smallest value in the half-open range [a, a+b) that is a multiple of b
3rd/cutlass/include/cutlass/fast_math.h:212
↓ 37 callersMethodto_nv_bfloat16
3rd/cutlass/include/cutlass/bfloat16.h:195
↓ 36 callersMethodadd_pointer_offset
Adds a pointer offset in units of Element
3rd/cutlass/include/cutlass/epilogue/threadblock/predicated_tile_iterator.h:959
↓ 36 callersFunctionfill_tma_gmem_shape_stride
3rd/cutlass/include/cute/atom/copy_traits_sm90_tma.hpp:855
↓ 36 callersMethodphase
3rd/cutlass/include/cutlass/pipeline/sm90_pipeline.hpp:193
↓ 35 callersFunctionproduct_each
3rd/cutlass/include/cute/int_tuple.hpp:247
↓ 35 callersFunctionstore_with_pointer_offset
Store a fragment to memory
3rd/cutlass/include/cutlass/transform/threadblock/regular_tile_iterator_tensor_op_sm70.h:237
↓ 35 callersFunctiontma_descriptor_replace_dims_strides_in_shared_mem
Replace tensor dims and strides for GEMMs by bringing the tensormap from GMEM into the shared memory
3rd/cutlass/include/cute/arch/copy_sm90_desc.hpp:358
↓ 34 callersFunctionsynclog_alloc
3rd/cutlass/include/cutlass/arch/synclog.hpp:65
↓ 34 callersFunctionsynclog_condition_emit
3rd/cutlass/include/cutlass/arch/synclog.hpp:226
↓ 34 callersFunctionsynclog_emit_prefix
3rd/cutlass/include/cutlass/arch/synclog.hpp:77
↓ 34 callersFunctionsynclog_print_prefix
3rd/cutlass/include/cutlass/arch/synclog.hpp:246
↓ 33 callersMethodadvance_to_next_work
3rd/cutlass/include/cutlass/gemm/kernel/sm100_tile_scheduler.hpp:439
↓ 33 callersFunctionblock_rank_in_cluster
Get 1D ctaid in a cluster.
3rd/cutlass/include/cute/arch/cluster_sm90.hpp:154
↓ 33 callersFunctionfence_view_async_shared
Issue a shared memory fence for async operations
3rd/cutlass/include/cutlass/arch/barrier.h:727
↓ 33 callersFunctionfind_tmem_tensor_col_offset
3rd/cutlass/include/cutlass/detail/sm100_tmem_helper.hpp:48
↓ 33 callersMethodinit
3rd/cutlass/include/cutlass/arch/barrier.h:355
↓ 33 callersMethodmax
Maximum finite value
3rd/cutlass/include/cutlass/float8.h:1356
↓ 33 callersMethodraw_data
3rd/cutlass/include/cutlass/array.h:415
↓ 32 callersFunctionallclose
(ref_tensor, real_tensor, atol=1e-8, rtol=1e-5)
tests/utils.py:176
↓ 31 callersMethodstore
Stores a fragment to memory at the location pointed to by the iterator
3rd/cutlass/include/cutlass/gemm/warp/mma_tensor_op_tile_iterator_wmma.h:777
↓ 30 callersFunctionTensorRef_aligned
3rd/cutlass/include/cutlass/tensor_ref.h:398
↓ 30 callersFunctioncomposition
3rd/cutlass/include/cute/layout_composed.hpp:361
↓ 30 callersMethoddivide
Alias for `div` to match the interface of FastDivmodU64
3rd/cutlass/include/cutlass/fast_math.h:396
↓ 30 callersFunctionget
Unpacks an element from memory
3rd/cutlass/include/cutlass/subbyte_reference.h:180
↓ 30 callersFunctionmake_coord_iterator
3rd/cutlass/include/cute/stride.hpp:556
↓ 29 callersMethodinit_masks
3rd/cutlass/include/cutlass/pipeline/sm100_pipeline.hpp:152
↓ 29 callersMethodis_producer_load_needed
3rd/cutlass/include/cutlass/epilogue/collective/detail.hpp:313
↓ 29 callersFunctionmul
3rd/cutlass/include/cute/numeric/real.hpp:49
← previousnext →101–200 of 10,542, ranked by callers