tilelang.contrib.cutedsl.warp ============================= .. py:module:: tilelang.contrib.cutedsl.warp .. autoapi-nested-parse:: Warp-level primitives for CuTeDSL backend. Re-exports from cutlass.cute.arch with TileLang naming conventions. Functions --------- .. autoapisummary:: tilelang.contrib.cutedsl.warp.warp_reduce_sum tilelang.contrib.cutedsl.warp.warp_reduce_max tilelang.contrib.cutedsl.warp.warp_reduce_min tilelang.contrib.cutedsl.warp.warp_reduce_bitand tilelang.contrib.cutedsl.warp.warp_reduce_bitor Module Contents --------------- .. py:function:: warp_reduce_sum(value) Warp-level parallel reduction: sum across all 32 lanes. .. py:function:: warp_reduce_max(value) Warp-level parallel reduction: max across all 32 lanes. .. py:function:: warp_reduce_min(value) Warp-level parallel reduction: min across all 32 lanes. .. py:function:: warp_reduce_bitand(value) Warp-level parallel reduction: bitwise AND across all 32 lanes. .. py:function:: warp_reduce_bitor(value) Warp-level parallel reduction: bitwise OR across all 32 lanes.