indirect_vreg_stream
|
indirect_offsets=(u6), valid_offset_mask=(u6), off_tile_start_offset_valid=(u1), off_tile_start_offset=(u5), sync_flag_core_type=(tec_scs / tac), sync_flag_count_type=(word_4b / descriptor), set_done_bit=(u1), tile_local_stride=(stride_32b / stride_64b / stride_128b / stride_256b / stride_512b / stride_1024b / stride_2048b / no_stride), post_update_circular_buffer=(u1), indirect_list_type=(word_offset / row_offset), indirect_list_stride=(u6), indirect_filter_en=(u1), trace_en=(u1), stream_opcode=(gather / gather_integer_add / gather_float_add / reserved_0 / scatter / scatter_integer_add / scatter_float_add / reserved_1), tile_local_memory_type=(smem / tile_spmem), tile_local_stream_type=(linear / circular_buffer), off_tile_memory_type=(spmem / tile_spmem_n / hbm / hbm_4b), gather_scatter_add_is_b16=(u1), s0_x=(u5), s0_y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32), s1_x=(u5), s1_y=(sreg0 / sreg1 / sreg2 / sreg3 / sreg4 / sreg5 / sreg6 / sreg7 / sreg8 / sreg9 / sreg10 / sreg11 / sreg12 / sreg13 / sreg14 / sreg15 / sreg16 / sreg17 / sreg18 / sreg19 / sreg20 / sreg21 / sreg22 / sreg23 / sreg24 / sreg25 / sreg26 / sreg27 / sreg28 / sreg29 / sreg30 / sreg31 / u32)
|
stream.indirect.gather
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.gather.add.bf16
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.gather.add.f32
|
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.indirect.gather.add.s16
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.gather.add.s32
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.gather.cb
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.gather.cb.add.bf16
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.gather.cb.add.f32
|
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.indirect.gather.cb.add.s16
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.gather.cb.add.s32
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter.add.bf16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter.add.f32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.indirect.scatter.add.s16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter.add.s32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter.cb
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter.cb.add.bf16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter.cb.add.f32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.indirect.scatter.cb.add.s16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.indirect.scatter.cb.add.s32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.linear.gather
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], u32
|
stream.linear.gather.add.bf16
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
|
stream.linear.gather.add.f32
|
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], f32, u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], f32, u32
|
stream.linear.gather.add.s16
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
|
stream.linear.gather.add.s32
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
|
stream.linear.gather.cb
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], u32
|
stream.linear.gather.cb.add.bf16
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
|
stream.linear.gather.cb.add.f32
|
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], f32, u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], f32, u32
|
stream.linear.gather.cb.add.s16
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
|
stream.linear.gather.cb.add.s32
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], u32
|
stream.linear.scatter
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
|
stream.linear.scatter.add.bf16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
|
stream.linear.scatter.add.f32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, u32
|
stream.linear.scatter.add.s16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
|
stream.linear.scatter.add.s32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], u32
|
stream.linear.scatter.cb
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
|
stream.linear.scatter.cb.add.bf16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
|
stream.linear.scatter.cb.add.f32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], f32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], f32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, u32
|
stream.linear.scatter.cb.add.s16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
|
stream.linear.scatter.cb.add.s32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], u32
|
stream.strided.gather
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.gather.add.bf16
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.gather.add.f32
|
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.strided.gather.add.s16
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.gather.add.s32
|
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:s[0..31]], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.gather.cb
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.gather.cb.add.bf16
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.gather.cb.add.f32
|
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:f32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], f32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.strided.gather.cb.add.s16
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.gather.cb.add.s32
|
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:u32], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], u32, s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [hbm4b:s[0..31]+s[0..31]], s[0..31], s[0..31], s[0..31], u32
[tilespmem:u5], [sflag:s[0..31]], [spmem:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter.add.bf16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter.add.f32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.strided.scatter.add.s16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter.add.s32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:s[0..31]], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter.cb
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter.cb.add.bf16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter.cb.add.f32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], f32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:f32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], f32, s[0..31], s[0..31], u32
|
stream.strided.scatter.cb.add.s16
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|
stream.strided.scatter.cb.add.s32
|
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:u32], s[0..31], s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], u32, s[0..31], s[0..31], u32
[hbm4b:s[0..31]+s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
[spmem:s[0..31]], [tilespmem:u5], [sflag:s[0..31]], s[0..31], s[0..31], s[0..31], u32
|