submission 368725

import torch
import cutlass
import cutlass.cute as cute
from cutlass.cute.nvgpu import cpasync, tcgen05
import cutlass.pipeline as pipeline
import cutlass.utils as utils
import cutlass.utils.blackwell_helpers as sm100_utils
import cutlass.utils.blockscaled_layout as blockscaled_utils
from cutlass.cute.runtime import make_ptr
import cutlass.cute.math as cute_math
from typing import Tuple, Type, Union
_SF_VEC_SIZE = 16
_MMA_TILER_MN = (128, 64)
_CLUSTER_SHAPE_MN = (1, 2)
_OCCUPANCY = 1
_TMA_CACHE_EVICT_NORMAL = 0x1000000000000000
_TMA_CACHE_EVICT_FIRST = 0x12F0000000000000
_TMA_CACHE_EVICT_LAST = 0x14F0000000000000
class Sm100BlockScaledPersistentDenseGemmKernel:
    def __init__(
        self,
        sf_vec_size: int,
        mma_tiler_mn: Tuple[int, int],
        cluster_shape_mn: Tuple[int, int],
        occupancy: int = 1,
    ):
        self.acc_dtype = cutlass.Float32
        self.sf_vec_size = sf_vec_size
        self.use_2cta_instrs = mma_tiler_mn[0] == 256
        self.cluster_shape_mn = cluster_shape_mn
        self.mma_tiler = (*mma_tiler_mn, 1)
        self.cta_group = (
            tcgen05.CtaGroup.TWO if self.use_2cta_instrs else tcgen05.CtaGroup.ONE
        )
        self.occupancy = int(occupancy)
        self.epilog_warp_id = (
            0,
            1,
            2,
            3,
        )
        self.mma_warp_id = 4
        self.tma_warp_id = 5
        self.threads_per_cta = 32 * len(
            (self.mma_warp_id, self.tma_warp_id, *self.epilog_warp_id)
        )
        self.epilog_sync_barrier = pipeline.NamedBarrier(
            barrier_id=1,
            num_threads=32 * len(self.epilog_warp_id),
        )
        self.tmem_alloc_barrier = pipeline.NamedBarrier(
            barrier_id=2,
            num_threads=32 * len((self.mma_warp_id, *self.epilog_warp_id)),
        )
        self.smem_capacity = utils.get_smem_capacity_in_bytes("sm_100")
        self.num_tmem_alloc_cols = 512
    def _setup_attributes(self):
        self.mma_inst_shape_mn = (
            self.mma_tiler[0],
            self.mma_tiler[1],
        )
        self.mma_inst_shape_mn_sfb = (
            self.mma_inst_shape_mn[0] // (2 if self.use_2cta_instrs else 1),
            cute.round_up(self.mma_inst_shape_mn[1], 128),
        )
        tiled_mma = sm100_utils.make_blockscaled_trivial_tiled_mma(
            self.a_dtype,
            self.a_major_mode,
            self.b_major_mode,
            self.sf_dtype,
            self.sf_vec_size,
            self.cta_group,
            self.mma_inst_shape_mn,
        )
        tiled_mma_sfb = sm100_utils.make_blockscaled_trivial_tiled_mma(
            self.a_dtype,
            self.a_major_mode,
            self.b_major_mode,
            self.sf_dtype,
            self.sf_vec_size,
            cute.nvgpu.tcgen05.CtaGroup.ONE,
            self.mma_inst_shape_mn_sfb,
        )
        mma_inst_shape_k = cute.size(tiled_mma.shape_mnk, mode=[2])
        mma_inst_tile_k = 4
        self.mma_tiler = (
            self.mma_inst_shape_mn[0],
            self.mma_inst_shape_mn[1],
            mma_inst_shape_k * mma_inst_tile_k,
        )
        self.mma_tiler_sfb = (
            self.mma_inst_shape_mn_sfb[0],
            self.mma_inst_shape_mn_sfb[1],
            mma_inst_shape_k * mma_inst_tile_k,
        )
        self.cta_tile_shape_mnk = (
            self.mma_tiler[0] // cute.size(tiled_mma.thr_id.shape),
            self.mma_tiler[1],
            self.mma_tiler[2],
        )
        self.cta_tile_shape_mnk_sfb = (
            self.mma_tiler_sfb[0] // cute.size(tiled_mma.thr_id.shape),
            self.mma_tiler_sfb[1],
            self.mma_tiler_sfb[2],
        )
        self.cluster_layout_vmnk = cute.tiled_divide(
            cute.make_layout((*self.cluster_shape_mn, 1)),
            (tiled_mma.thr_id.shape,),
        )
        self.cluster_layout_sfb_vmnk = cute.tiled_divide(
            cute.make_layout((*self.cluster_shape_mn, 1)),
            (tiled_mma_sfb.thr_id.shape,),
        )
        self.num_mcast_ctas_a = cute.size(self.cluster_layout_vmnk.shape[2])
        self.num_mcast_ctas_b = cute.size(self.cluster_layout_vmnk.shape[1])
        self.num_mcast_ctas_sfb = cute.size(self.cluster_layout_sfb_vmnk.shape[1])
        self.is_a_mcast = self.num_mcast_ctas_a > 1
        self.is_b_mcast = self.num_mcast_ctas_b > 1
        self.is_sfb_mcast = self.num_mcast_ctas_sfb > 1
        self.epi_tile = sm100_utils.compute_epilogue_tile_shape(
            self.cta_tile_shape_mnk,
            self.use_2cta_instrs,
            self.c_layout,
            self.c_dtype,
        )
        self.epi_tile_n = cute.size(self.epi_tile[1])
        self.num_acc_stage, self.num_ab_stage, self.num_c_stage = self._compute_stages(
            tiled_mma,
            self.mma_tiler,
            self.a_dtype,
            self.b_dtype,
            self.epi_tile,
            self.c_dtype,
            self.c_layout,
            self.sf_dtype,
            self.sf_vec_size,
            self.smem_capacity,
            self.occupancy,
        )
        self.a_smem_layout_staged = sm100_utils.make_smem_layout_a(
            tiled_mma,
            self.mma_tiler,
            self.a_dtype,
            self.num_ab_stage,
        )
        self.b_smem_layout_staged = sm100_utils.make_smem_layout_b(
            tiled_mma,
            self.mma_tiler,
            self.b_dtype,
            self.num_ab_stage,
        )
        self.sfa_smem_layout_staged = blockscaled_utils.make_smem_layout_sfa(
            tiled_mma,
            self.mma_tiler,
            self.sf_vec_size,
            self.num_ab_stage,
        )
        self.sfb_smem_layout_staged = blockscaled_utils.make_smem_layout_sfb(
            tiled_mma,
            self.mma_tiler,
            self.sf_vec_size,
            self.num_ab_stage,
        )
        self.c_smem_layout_staged = sm100_utils.make_smem_layout_epi(
            self.c_dtype,
            self.c_layout,
            self.epi_tile,
            self.num_c_stage,
        )
        self.overlapping_accum = self.num_acc_stage == 1
        sf_atom_mn = 32
        self.num_sfa_tmem_cols = (self.cta_tile_shape_mnk[0] // sf_atom_mn) * mma_inst_tile_k
        self.num_sfb_tmem_cols = (self.cta_tile_shape_mnk_sfb[1] // sf_atom_mn) * mma_inst_tile_k
        # Fused dual GEMM: need space for both SFB1 and SFB2
        self.num_sf_tmem_cols = self.num_sfa_tmem_cols + self.num_sfb_tmem_cols * 2
        # For dual GEMM, we need to store two accumulators
        self.num_accumulator_tmem_cols_single = self.cta_tile_shape_mnk[1] * self.num_acc_stage if not self.overlapping_accum else self.cta_tile_shape_mnk[1] * 2 - self.num_sf_tmem_cols
        self.num_accumulator_tmem_cols = self.num_accumulator_tmem_cols_single * 2  # Double for dual GEMM
        self.iter_acc_early_release_in_epilogue = self.num_sf_tmem_cols // self.epi_tile_n
    @cute.jit
    def __call__(
        self,
        a_ptr: cute.Pointer,
        b1_ptr: cute.Pointer,
        b2_ptr: cute.Pointer,
        sfa_ptr: cute.Pointer,
        sfb1_ptr: cute.Pointer,
        sfb2_ptr: cute.Pointer,
        c_ptr: cute.Pointer,
        problem_size: tuple,
        max_active_clusters: cutlass.Constexpr,
    ):
        m, n, k, l = problem_size
        sf_k = k // self.sf_vec_size
        a_tensor = cute.make_tensor(
            a_ptr,
            cute.make_layout(
                (m, cute.assume(k, 32), l),
                stride=(cute.assume(k, 32), 1, cute.assume(m * k, 32)),
            ),
        )
        b1_tensor = cute.make_tensor(
            b1_ptr,
            cute.make_layout(
                (n, cute.assume(k, 32), l),
                stride=(cute.assume(k, 32), 1, cute.assume(n * k, 32)),
            ),
        )
        b2_tensor = cute.make_tensor(
            b2_ptr,
            cute.make_layout(
                (n, cute.assume(k, 32), l),
                stride=(cute.assume(k, 32), 1, cute.assume(n * k, 32)),
            ),
        )
        c_tensor = cute.make_tensor(
            c_ptr,
            cute.make_layout((cute.assume(m, 32), n, l), stride=(n, 1, m * n)),
        )
        sfa_tensor = cute.make_tensor(
            sfa_ptr,
            cute.make_layout((m, sf_k, l), stride=(sf_k, 1, m * sf_k)),
        )
        sfb1_tensor = cute.make_tensor(
            sfb1_ptr,
            cute.make_layout((n, sf_k, l), stride=(sf_k, 1, n * sf_k)),
        )
        sfb2_tensor = cute.make_tensor(
            sfb2_ptr,
            cute.make_layout((n, sf_k, l), stride=(sf_k, 1, n * sf_k)),
        )
        a_tensor.mark_compact_shape_dynamic(mode=1, stride_order=(2, 0, 1), divisibility=2)
        b1_tensor.mark_compact_shape_dynamic(mode=1, stride_order=(2, 0, 1), divisibility=2)
        b2_tensor.mark_compact_shape_dynamic(mode=1, stride_order=(2, 0, 1), divisibility=2)
        c_tensor.mark_compact_shape_dynamic(mode=1, stride_order=(2, 0, 1), divisibility=1)
        self.a_dtype: Type[cutlass.Numeric] = a_tensor.element_type
        self.b_dtype: Type[cutlass.Numeric] = b1_tensor.element_type
        self.sf_dtype: Type[cutlass.Numeric] = sfa_tensor.element_type
        self.c_dtype: Type[cutlass.Numeric] = c_tensor.element_type
        self.a_major_mode = utils.LayoutEnum.from_tensor(a_tensor).mma_major_mode()
        self.b_major_mode = utils.LayoutEnum.from_tensor(b1_tensor).mma_major_mode()
        self.c_layout = utils.LayoutEnum.from_tensor(c_tensor)
        if cutlass.const_expr(self.a_dtype != self.b_dtype):
            raise TypeError(f"Type must match: {self.a_dtype} != {self.b_dtype}")
        self._setup_attributes()
        sfa_layout = blockscaled_utils.tile_atom_to_shape_SF(
            a_tensor.shape, self.sf_vec_size
        )
        sfa_tensor = cute.make_tensor(sfa_tensor.iterator, sfa_layout)
        sfb1_layout = blockscaled_utils.tile_atom_to_shape_SF(
            b1_tensor.shape, self.sf_vec_size
        )
        sfb1_tensor = cute.make_tensor(sfb1_tensor.iterator, sfb1_layout)
        sfb2_layout = blockscaled_utils.tile_atom_to_shape_SF(
            b2_tensor.shape, self.sf_vec_size
        )
        sfb2_tensor = cute.make_tensor(sfb2_tensor.iterator, sfb2_layout)
        tiled_mma = sm100_utils.make_blockscaled_trivial_tiled_mma(
            self.a_dtype,
            self.a_major_mode,
            self.b_major_mode,
            self.sf_dtype,
            self.sf_vec_size,
            self.cta_group,
            self.mma_inst_shape_mn,
        )
        tiled_mma_sfb = sm100_utils.make_blockscaled_trivial_tiled_mma(
            self.a_dtype,
            self.a_major_mode,
            self.b_major_mode,
            self.sf_dtype,
            self.sf_vec_size,
            cute.nvgpu.tcgen05.CtaGroup.ONE,
            self.mma_inst_shape_mn_sfb,
        )
        atom_thr_size = cute.size(tiled_mma.thr_id.shape)
        a_op = sm100_utils.cluster_shape_to_tma_atom_A(
            self.cluster_shape_mn, tiled_mma.thr_id
        )
        a_smem_layout = cute.slice_(self.a_smem_layout_staged, (None, None, None, 0))
        tma_atom_a, tma_tensor_a = cute.nvgpu.make_tiled_tma_atom_A(
            a_op,
            a_tensor,
            a_smem_layout,
            self.mma_tiler,
            tiled_mma,
            self.cluster_layout_vmnk.shape,
        )
        b_op = sm100_utils.cluster_shape_to_tma_atom_B(
            self.cluster_shape_mn, tiled_mma.thr_id
        )
        b_smem_layout = cute.slice_(self.b_smem_layout_staged, (None, None, None, 0))
        tma_atom_b1, tma_tensor_b1 = cute.nvgpu.make_tiled_tma_atom_B(
            b_op,
            b1_tensor,
            b_smem_layout,
            self.mma_tiler,
            tiled_mma,
            self.cluster_layout_vmnk.shape,
        )
        tma_atom_b2, tma_tensor_b2 = cute.nvgpu.make_tiled_tma_atom_B(
            b_op,
            b2_tensor,
            b_smem_layout,
            self.mma_tiler,
            tiled_mma,
            self.cluster_layout_vmnk.shape,
        )
        sfa_op = sm100_utils.cluster_shape_to_tma_atom_A(
            self.cluster_shape_mn, tiled_mma.thr_id
        )
        sfa_smem_layout = cute.slice_(
            self.sfa_smem_layout_staged, (None, None, None, 0)
        )
        tma_atom_sfa, tma_tensor_sfa = cute.nvgpu.make_tiled_tma_atom_A(
            sfa_op,
            sfa_tensor,
            sfa_smem_layout,
            self.mma_tiler,
            tiled_mma,
            self.cluster_layout_vmnk.shape,
            internal_type=cutlass.Int16,
        )
        sfb_op = sm100_utils.cluster_shape_to_tma_atom_SFB(
            self.cluster_shape_mn, tiled_mma.thr_id
        )
        sfb_smem_layout = cute.slice_(
            self.sfb_smem_layout_staged, (None, None, None, 0)
        )
        tma_atom_sfb1, tma_tensor_sfb1 = cute.nvgpu.make_tiled_tma_atom_B(
            sfb_op,
            sfb1_tensor,
            sfb_smem_layout,
            self.mma_tiler_sfb,
            tiled_mma_sfb,
            self.cluster_layout_sfb_vmnk.shape,
            internal_type=cutlass.Int16,
        )
        tma_atom_sfb2, tma_tensor_sfb2 = cute.nvgpu.make_tiled_tma_atom_B(
            sfb_op,
            sfb2_tensor,
            sfb_smem_layout,
            self.mma_tiler_sfb,
            tiled_mma_sfb,
            self.cluster_layout_sfb_vmnk.shape,
            internal_type=cutlass.Int16,
        )
        if cutlass.const_expr(self.cta_tile_shape_mnk[1] == 192):
            x = tma_tensor_sfb1.stride[0][1]
            y = cute.ceil_div(tma_tensor_sfb1.shape[0][1], 4)
            new_shape = (
                (
                    tma_tensor_sfb1.shape[0][0],
                    ((2, 2), y)
                ),
                tma_tensor_sfb1.shape[1],
                tma_tensor_sfb1.shape[2]
            )
            x_times_3 = 3 * x
            new_stride = (
                (
                    tma_tensor_sfb1.stride[0][0],
                    ((x, x), x_times_3)
                ),
                tma_tensor_sfb1.stride[1],
                tma_tensor_sfb1.stride[2]
            )
            tma_tensor_sfb_new_layout = cute.make_layout(new_shape, stride=new_stride)
            tma_tensor_sfb1 = cute.make_tensor(tma_tensor_sfb1.iterator, tma_tensor_sfb_new_layout)
            tma_tensor_sfb2 = cute.make_tensor(tma_tensor_sfb2.iterator, tma_tensor_sfb_new_layout)
        a_copy_size = cute.size_in_bytes(self.a_dtype, a_smem_layout)
        b_copy_size = cute.size_in_bytes(self.b_dtype, b_smem_layout)
        sfa_copy_size = cute.size_in_bytes(self.sf_dtype, sfa_smem_layout)
        sfb_copy_size = cute.size_in_bytes(self.sf_dtype, sfb_smem_layout)
        # Fused dual GEMM: load A, B1, B2, SFA, SFB1, SFB2 in one iteration
        self.num_tma_load_bytes = (
            a_copy_size + b_copy_size * 2 + sfa_copy_size + sfb_copy_size * 2
        ) * atom_thr_size
        epi_smem_layout = cute.slice_(self.c_smem_layout_staged, (None, None, 0))
        tma_atom_c, tma_tensor_c = cpasync.make_tiled_tma_atom(
            cpasync.CopyBulkTensorTileS2GOp(),
            c_tensor,
            epi_smem_layout,
            self.epi_tile,
        )
        grid_m = (m // self.cta_tile_shape_mnk[0]) * cute.size(tiled_mma.thr_id.shape)
        grid_n = n // self.cta_tile_shape_mnk[1]
        grid = (grid_m, grid_n, l)
        self.buffer_align_bytes = 1024
        @cute.struct
        class SharedStorage:
            ab_full_mbar_ptr: cute.struct.MemRange[cutlass.Int64, self.num_ab_stage]
            ab_empty_mbar_ptr: cute.struct.MemRange[cutlass.Int64, self.num_ab_stage]
            acc_full_mbar_ptr: cute.struct.MemRange[cutlass.Int64, self.num_acc_stage]
            acc_empty_mbar_ptr: cute.struct.MemRange[cutlass.Int64, self.num_acc_stage]
            tmem_dealloc_mbar_ptr: cutlass.Int64
            tmem_holding_buf: cutlass.Int32
            sC: cute.struct.Align[
                cute.struct.MemRange[
                    self.c_dtype,
                    cute.cosize(self.c_smem_layout_staged.outer),
                ],
                self.buffer_align_bytes,
            ]
            sA: cute.struct.Align[
                cute.struct.MemRange[
                    self.a_dtype, cute.cosize(self.a_smem_layout_staged.outer)
                ],
                self.buffer_align_bytes,
            ]
            sB1: cute.struct.Align[
                cute.struct.MemRange[
                    self.b_dtype, cute.cosize(self.b_smem_layout_staged.outer)
                ],
                self.buffer_align_bytes,
            ]
            sB2: cute.struct.Align[
                cute.struct.MemRange[
                    self.b_dtype, cute.cosize(self.b_smem_layout_staged.outer)
                ],
                self.buffer_align_bytes,
            ]
            sSFA: cute.struct.Align[
                cute.struct.MemRange[
                    self.sf_dtype, cute.cosize(self.sfa_smem_layout_staged)
                ],
                self.buffer_align_bytes,
            ]
            sSFB1: cute.struct.Align[
                cute.struct.MemRange[
                    self.sf_dtype, cute.cosize(self.sfb_smem_layout_staged)
                ],
                self.buffer_align_bytes,
            ]
            sSFB2: cute.struct.Align[
                cute.struct.MemRange[
                    self.sf_dtype, cute.cosize(self.sfb_smem_layout_staged)
                ],
                self.buffer_align_bytes,
            ]
        self.shared_storage = SharedStorage
        self.kernel(
            tiled_mma,
            tiled_mma_sfb,
            tma_atom_a,
            tma_tensor_a,
            tma_atom_b1,
            tma_tensor_b1,
            tma_atom_b2,
            tma_tensor_b2,
            tma_atom_sfa,
            tma_tensor_sfa,
            tma_atom_sfb1,
            tma_tensor_sfb1,
            tma_atom_sfb2,
            tma_tensor_sfb2,
            tma_atom_c,
            tma_tensor_c,
            self.cluster_layout_vmnk,
            self.cluster_layout_sfb_vmnk,
            self.a_smem_layout_staged,
            self.b_smem_layout_staged,
            self.sfa_smem_layout_staged,
            self.sfb_smem_layout_staged,
            self.c_smem_layout_staged,
            self.epi_tile,
        ).launch(
            grid=grid,
            block=[self.threads_per_cta, 1, 1],
            cluster=(*self.cluster_shape_mn, 1),
            min_blocks_per_mp=self.occupancy,
        )
        return
    @cute.kernel
    def kernel(
        self,
        tiled_mma: cute.TiledMma,
        tiled_mma_sfb: cute.TiledMma,
        tma_atom_a: cute.CopyAtom,
        mA_mkl: cute.Tensor,
        tma_atom_b1: cute.CopyAtom,
        mB1_nkl: cute.Tensor,
        tma_atom_b2: cute.CopyAtom,
        mB2_nkl: cute.Tensor,
        tma_atom_sfa: cute.CopyAtom,
        mSFA_mkl: cute.Tensor,
        tma_atom_sfb1: cute.CopyAtom,
        mSFB1_nkl: cute.Tensor,
        tma_atom_sfb2: cute.CopyAtom,
        mSFB2_nkl: cute.Tensor,
        tma_atom_c: cute.CopyAtom,
        mC_mnl: cute.Tensor,
        cluster_layout_vmnk: cute.Layout,
        cluster_layout_sfb_vmnk: cute.Layout,
        a_smem_layout_staged: cute.ComposedLayout,
        b_smem_layout_staged: cute.ComposedLayout,
        sfa_smem_layout_staged: cute.Layout,
        sfb_smem_layout_staged: cute.Layout,
        c_smem_layout_staged: Union[cute.Layout, cute.ComposedLayout],
        epi_tile: cute.Tile,
    ):
        warp_idx = cute.arch.warp_idx()
        warp_idx = cute.arch.make_warp_uniform(warp_idx)
        if warp_idx == self.tma_warp_id:
            cpasync.prefetch_descriptor(tma_atom_a)
            cpasync.prefetch_descriptor(tma_atom_b1)
            cpasync.prefetch_descriptor(tma_atom_b2)
            cpasync.prefetch_descriptor(tma_atom_sfa)
            cpasync.prefetch_descriptor(tma_atom_sfb1)
            cpasync.prefetch_descriptor(tma_atom_sfb2)
            cpasync.prefetch_descriptor(tma_atom_c)
        use_2cta_instrs = cute.size(tiled_mma.thr_id.shape) == 2
        bidx, bidy, bidz = cute.arch.block_idx()
        mma_tile_coord_v = bidx % cute.size(tiled_mma.thr_id.shape)
        is_leader_cta = mma_tile_coord_v == 0
        cta_rank_in_cluster = cute.arch.make_warp_uniform(
            cute.arch.block_idx_in_cluster()
        )
        block_in_cluster_coord_vmnk = cluster_layout_vmnk.get_flat_coord(
            cta_rank_in_cluster
        )
        block_in_cluster_coord_sfb_vmnk = cluster_layout_sfb_vmnk.get_flat_coord(
            cta_rank_in_cluster
        )
        tidx, _, _ = cute.arch.thread_idx()
        smem = utils.SmemAllocator()
        storage = smem.allocate(self.shared_storage)
        ab_pipeline_producer_group = pipeline.CooperativeGroup(pipeline.Agent.Thread)
        num_tma_producer = self.num_mcast_ctas_a + self.num_mcast_ctas_b - 1
        ab_pipeline_consumer_group = pipeline.CooperativeGroup(
            pipeline.Agent.Thread, num_tma_producer
        )
        try:
            ab_pipeline = pipeline.PipelineTmaUmma.create(
                barrier_storage=storage.ab_full_mbar_ptr.data_ptr(),
                num_stages=self.num_ab_stage,
                producer_group=ab_pipeline_producer_group,
                consumer_group=ab_pipeline_consumer_group,
                tx_count=self.num_tma_load_bytes,
                cta_layout_vmnk=cluster_layout_vmnk,
                defer_sync=True,
            )
        except TypeError:
            ab_pipeline = pipeline.PipelineTmaUmma.create(
                barrier_storage=storage.ab_full_mbar_ptr.data_ptr(),
                num_stages=self.num_ab_stage,
                producer_group=ab_pipeline_producer_group,
                consumer_group=ab_pipeline_consumer_group,
                tx_count=self.num_tma_load_bytes,
                cta_layout_vmnk=cluster_layout_vmnk,
            )
        acc_pipeline_producer_group = pipeline.CooperativeGroup(pipeline.Agent.Thread)
        num_acc_consumer_threads = len(self.epilog_warp_id) * (
            2 if use_2cta_instrs else 1
        )
        acc_pipeline_consumer_group = pipeline.CooperativeGroup(
            pipeline.Agent.Thread, num_acc_consumer_threads
        )
        try:
            acc_pipeline = pipeline.PipelineUmmaAsync.create(
                barrier_storage=storage.acc_full_mbar_ptr.data_ptr(),
                num_stages=self.num_acc_stage,
                producer_group=acc_pipeline_producer_group,
                consumer_group=acc_pipeline_consumer_group,
                cta_layout_vmnk=cluster_layout_vmnk,
                defer_sync=True,
            )
        except TypeError:
            acc_pipeline = pipeline.PipelineUmmaAsync.create(
                barrier_storage=storage.acc_full_mbar_ptr.data_ptr(),
                num_stages=self.num_acc_stage,
                producer_group=acc_pipeline_producer_group,
                consumer_group=acc_pipeline_consumer_group,
                cta_layout_vmnk=cluster_layout_vmnk,
            )
        tmem = utils.TmemAllocator(
            storage.tmem_holding_buf,
            barrier_for_retrieve=self.tmem_alloc_barrier,
            allocator_warp_id=self.epilog_warp_id[0],
            is_two_cta=use_2cta_instrs,
            two_cta_tmem_dealloc_mbar_ptr=storage.tmem_dealloc_mbar_ptr,
        )
        sC = storage.sC.get_tensor(
            c_smem_layout_staged.outer, swizzle=c_smem_layout_staged.inner
        )
        sA = storage.sA.get_tensor(
            a_smem_layout_staged.outer, swizzle=a_smem_layout_staged.inner
        )
        sB1 = storage.sB1.get_tensor(
            b_smem_layout_staged.outer, swizzle=b_smem_layout_staged.inner
        )
        sB2 = storage.sB2.get_tensor(
            b_smem_layout_staged.outer, swizzle=b_smem_layout_staged.inner
        )
        sSFA = storage.sSFA.get_tensor(sfa_smem_layout_staged)
        sSFB1 = storage.sSFB1.get_tensor(sfb_smem_layout_staged)
        sSFB2 = storage.sSFB2.get_tensor(sfb_smem_layout_staged)
        a_full_mcast_mask = None
        b_full_mcast_mask = None
        sfa_full_mcast_mask = None
        sfb_full_mcast_mask = None
        if cutlass.const_expr(self.is_a_mcast or self.is_b_mcast or use_2cta_instrs):
            a_full_mcast_mask = cpasync.create_tma_multicast_mask(
                cluster_layout_vmnk, block_in_cluster_coord_vmnk, mcast_mode=2
            )
            b_full_mcast_mask = cpasync.create_tma_multicast_mask(
                cluster_layout_vmnk, block_in_cluster_coord_vmnk, mcast_mode=1
            )
            sfa_full_mcast_mask = cpasync.create_tma_multicast_mask(
                cluster_layout_vmnk, block_in_cluster_coord_vmnk, mcast_mode=2
            )
            sfb_full_mcast_mask = cpasync.create_tma_multicast_mask(
                cluster_layout_sfb_vmnk, block_in_cluster_coord_sfb_vmnk, mcast_mode=1
            )
        gA_mkl = cute.local_tile(
            mA_mkl, cute.slice_(self.mma_tiler, (None, 0, None)), (None, None, None)
        )
        gB1_nkl = cute.local_tile(
            mB1_nkl, cute.slice_(self.mma_tiler, (0, None, None)), (None, None, None)
        )
        gB2_nkl = cute.local_tile(
            mB2_nkl, cute.slice_(self.mma_tiler, (0, None, None)), (None, None, None)
        )
        gSFA_mkl = cute.local_tile(
            mSFA_mkl, cute.slice_(self.mma_tiler, (None, 0, None)), (None, None, None)
        )
        gSFB1_nkl = cute.local_tile(
            mSFB1_nkl,
            cute.slice_(self.mma_tiler_sfb, (0, None, None)),
            (None, None, None),
        )
        gSFB2_nkl = cute.local_tile(
            mSFB2_nkl,
            cute.slice_(self.mma_tiler_sfb, (0, None, None)),
            (None, None, None),
        )
        gC_mnl = cute.local_tile(
            mC_mnl, cute.slice_(self.mma_tiler, (None, None, 0)), (None, None, None)
        )
        k_tile_cnt = cute.size(gA_mkl, mode=[3])
        thr_mma = tiled_mma.get_slice(mma_tile_coord_v)
        thr_mma_sfb = tiled_mma_sfb.get_slice(mma_tile_coord_v)
        tCgA = thr_mma.partition_A(gA_mkl)
        tCgB1 = thr_mma.partition_B(gB1_nkl)
        tCgB2 = thr_mma.partition_B(gB2_nkl)
        tCgSFA = thr_mma.partition_A(gSFA_mkl)
        tCgSFB1 = thr_mma_sfb.partition_B(gSFB1_nkl)
        tCgSFB2 = thr_mma_sfb.partition_B(gSFB2_nkl)
        tCgC = thr_mma.partition_C(gC_mnl)
        a_cta_layout = cute.make_layout(
            cute.slice_(cluster_layout_vmnk, (0, 0, None, 0)).shape
        )
        tAsA, tAgA = cpasync.tma_partition(
            tma_atom_a,
            block_in_cluster_coord_vmnk[2],
            a_cta_layout,
            cute.group_modes(sA, 0, 3),
            cute.group_modes(tCgA, 0, 3),
        )
        b_cta_layout = cute.make_layout(
            cute.slice_(cluster_layout_vmnk, (0, None, 0, 0)).shape
        )
        tBsB1, tBgB1 = cpasync.tma_partition(
            tma_atom_b1,
            block_in_cluster_coord_vmnk[1],
            b_cta_layout,
            cute.group_modes(sB1, 0, 3),
            cute.group_modes(tCgB1, 0, 3),
        )
        tBsB2, tBgB2 = cpasync.tma_partition(
            tma_atom_b2,
            block_in_cluster_coord_vmnk[1],
            b_cta_layout,
            cute.group_modes(sB2, 0, 3),
            cute.group_modes(tCgB2, 0, 3),
        )
        sfa_cta_layout = a_cta_layout
        tAsSFA, tAgSFA = cute.nvgpu.cpasync.tma_partition(
            tma_atom_sfa,
            block_in_cluster_coord_vmnk[2],
            sfa_cta_layout,
            cute.group_modes(sSFA, 0, 3),
            cute.group_modes(tCgSFA, 0, 3),
        )
        tAsSFA = cute.filter_zeros(tAsSFA)
        tAgSFA = cute.filter_zeros(tAgSFA)
        sfb_cta_layout = cute.make_layout(
            cute.slice_(cluster_layout_sfb_vmnk, (0, None, 0, 0)).shape
        )
        tBsSFB1, tBgSFB1 = cute.nvgpu.cpasync.tma_partition(
            tma_atom_sfb1,
            block_in_cluster_coord_sfb_vmnk[1],
            sfb_cta_layout,
            cute.group_modes(sSFB1, 0, 3),
            cute.group_modes(tCgSFB1, 0, 3),
        )
        tBsSFB1 = cute.filter_zeros(tBsSFB1)
        tBgSFB1 = cute.filter_zeros(tBgSFB1)
        tBsSFB2, tBgSFB2 = cute.nvgpu.cpasync.tma_partition(
            tma_atom_sfb2,
            block_in_cluster_coord_sfb_vmnk[1],
            sfb_cta_layout,
            cute.group_modes(sSFB2, 0, 3),
            cute.group_modes(tCgSFB2, 0, 3),
        )
        tBsSFB2 = cute.filter_zeros(tBsSFB2)
        tBgSFB2 = cute.filter_zeros(tBgSFB2)
        tCrA = tiled_mma.make_fragment_A(sA)
        tCrB1 = tiled_mma.make_fragment_B(sB1)
        tCrB2 = tiled_mma.make_fragment_B(sB2)
        acc_shape = tiled_mma.partition_shape_C(self.mma_tiler[:2])
        if cutlass.const_expr(self.overlapping_accum):
            num_acc_stage_overlapped = 2
            tCtAcc_fake = tiled_mma.make_fragment_C(
                cute.append(acc_shape, num_acc_stage_overlapped)
            )
            tCtAcc_fake = cute.make_tensor(
                tCtAcc_fake.iterator,
                cute.make_layout(
                    tCtAcc_fake.shape,
                    stride = (
                        tCtAcc_fake.stride[0],
                        tCtAcc_fake.stride[1],
                        tCtAcc_fake.stride[2],
                        (256 - self.num_sf_tmem_cols) * tCtAcc_fake.stride[0][1]
                    ) 
                )
            )
        else:
            tCtAcc_fake = tiled_mma.make_fragment_C(
                cute.append(acc_shape, self.num_acc_stage)
            )
        if warp_idx == self.tma_warp_id:
            ab_producer_state = pipeline.make_pipeline_state(
                pipeline.PipelineUserType.Producer, self.num_ab_stage
            )
            mma_tile_coord_mnl = (
                bidx // cute.size(tiled_mma.thr_id.shape),
                bidy,
                bidz,
            )
            tAgA_slice = tAgA[(None, mma_tile_coord_mnl[0], None, mma_tile_coord_mnl[2])]
            tBgB1_slice = tBgB1[(None, mma_tile_coord_mnl[1], None, mma_tile_coord_mnl[2])]
            tBgB2_slice = tBgB2[(None, mma_tile_coord_mnl[1], None, mma_tile_coord_mnl[2])]
            tAgSFA_slice = tAgSFA[(None, mma_tile_coord_mnl[0], None, mma_tile_coord_mnl[2])]
            slice_n = mma_tile_coord_mnl[1]
            if cutlass.const_expr(self.cta_tile_shape_mnk[1] == 64):
                slice_n = mma_tile_coord_mnl[1] // 2
            tBgSFB1_slice = tBgSFB1[(None, slice_n, None, mma_tile_coord_mnl[2])]
            tBgSFB2_slice = tBgSFB2[(None, slice_n, None, mma_tile_coord_mnl[2])]
            # Fused dual GEMM: load A, B1, B2, SFA, SFB1, SFB2 all per k-tile
            # This halves iteration count and allows both GEMMs per iteration
            total_k_tiles = k_tile_cnt
            ab_producer_state.reset_count()
            peek_ab_empty_status = cutlass.Boolean(1)
            if ab_producer_state.count < total_k_tiles:
                peek_ab_empty_status = ab_pipeline.producer_try_acquire(ab_producer_state)
            for k_tile in cutlass.range(0, total_k_tiles, 1, unroll=1):
                ab_pipeline.producer_acquire(ab_producer_state, peek_ab_empty_status)
                # Load A
                try:
                    cute.copy(
                        tma_atom_a,
                        tAgA_slice[(None, k_tile)],
                        tAsA[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=a_full_mcast_mask,
                        cache_policy=cutlass.Int64(cutlass.Int64(_TMA_CACHE_EVICT_FIRST).ir_value()),
                    )
                except TypeError:
                    cute.copy(
                        tma_atom_a,
                        tAgA_slice[(None, k_tile)],
                        tAsA[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=a_full_mcast_mask,
                    )
                # Load B1
                try:
                    cute.copy(
                        tma_atom_b1,
                        tBgB1_slice[(None, k_tile)],
                        tBsB1[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=b_full_mcast_mask,
                        cache_policy=cutlass.Int64(cutlass.Int64(_TMA_CACHE_EVICT_FIRST).ir_value()),
                    )
                except TypeError:
                    cute.copy(
                        tma_atom_b1,
                        tBgB1_slice[(None, k_tile)],
                        tBsB1[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=b_full_mcast_mask,
                    )
                # Load B2
                try:
                    cute.copy(
                        tma_atom_b2,
                        tBgB2_slice[(None, k_tile)],
                        tBsB2[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=b_full_mcast_mask,
                        cache_policy=cutlass.Int64(cutlass.Int64(_TMA_CACHE_EVICT_FIRST).ir_value()),
                    )
                except TypeError:
                    cute.copy(
                        tma_atom_b2,
                        tBgB2_slice[(None, k_tile)],
                        tBsB2[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=b_full_mcast_mask,
                    )
                # Load SFA
                try:
                    cute.copy(
                        tma_atom_sfa,
                        tAgSFA_slice[(None, k_tile)],
                        tAsSFA[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=sfa_full_mcast_mask,
                        cache_policy=cutlass.Int64(cutlass.Int64(_TMA_CACHE_EVICT_FIRST).ir_value()),
                    )
                except TypeError:
                    cute.copy(
                        tma_atom_sfa,
                        tAgSFA_slice[(None, k_tile)],
                        tAsSFA[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=sfa_full_mcast_mask,
                    )
                # Load SFB1
                try:
                    cute.copy(
                        tma_atom_sfb1,
                        tBgSFB1_slice[(None, k_tile)],
                        tBsSFB1[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=sfb_full_mcast_mask,
                        cache_policy=cutlass.Int64(cutlass.Int64(_TMA_CACHE_EVICT_FIRST).ir_value()),
                    )
                except TypeError:
                    cute.copy(
                        tma_atom_sfb1,
                        tBgSFB1_slice[(None, k_tile)],
                        tBsSFB1[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=sfb_full_mcast_mask,
                    )
                # Load SFB2
                try:
                    cute.copy(
                        tma_atom_sfb2,
                        tBgSFB2_slice[(None, k_tile)],
                        tBsSFB2[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=sfb_full_mcast_mask,
                        cache_policy=cutlass.Int64(cutlass.Int64(_TMA_CACHE_EVICT_FIRST).ir_value()),
                    )
                except TypeError:
                    cute.copy(
                        tma_atom_sfb2,
                        tBgSFB2_slice[(None, k_tile)],
                        tBsSFB2[(None, ab_producer_state.index)],
                        tma_bar_ptr=ab_pipeline.producer_get_barrier(ab_producer_state),
                        mcast_mask=sfb_full_mcast_mask,
                    )
                ab_producer_state.advance()
                peek_ab_empty_status = cutlass.Boolean(1)
                if ab_producer_state.count < total_k_tiles:
                    peek_ab_empty_status = ab_pipeline.producer_try_acquire(ab_producer_state)
            ab_pipeline.producer_tail(ab_producer_state)
        if warp_idx == self.mma_warp_id:
            tmem.wait_for_alloc()
            acc_tmem_ptr = tmem.retrieve_ptr(self.acc_dtype)
            # Create two accumulator bases - one for each GEMM
            tCtAcc1_base = cute.make_tensor(acc_tmem_ptr, tCtAcc_fake.layout)
            # Second accumulator offset by single accumulator size
            acc2_tmem_ptr = acc_tmem_ptr + self.num_accumulator_tmem_cols_single
            tCtAcc2_base = cute.make_tensor(acc2_tmem_ptr, tCtAcc_fake.layout)
            sfa_tmem_ptr = cute.recast_ptr(
                acc_tmem_ptr + self.num_accumulator_tmem_cols,
                dtype=self.sf_dtype,
            )
            tCtSFA_layout = blockscaled_utils.make_tmem_layout_sfa(
                tiled_mma,
                self.mma_tiler,
                self.sf_vec_size,
                cute.slice_(sfa_smem_layout_staged, (None, None, None, 0)),
            )
            tCtSFA = cute.make_tensor(sfa_tmem_ptr, tCtSFA_layout)
            sfb1_tmem_ptr = cute.recast_ptr(
                acc_tmem_ptr + self.num_accumulator_tmem_cols + self.num_sfa_tmem_cols,
                dtype=self.sf_dtype,
            )
            tCtSFB_layout = blockscaled_utils.make_tmem_layout_sfb(
                tiled_mma,
                self.mma_tiler,
                self.sf_vec_size,
                cute.slice_(sfb_smem_layout_staged, (None, None, None, 0)),
            )
            tCtSFB1 = cute.make_tensor(sfb1_tmem_ptr, tCtSFB_layout)
            sfb2_tmem_ptr = cute.recast_ptr(
                acc_tmem_ptr + self.num_accumulator_tmem_cols + self.num_sfa_tmem_cols + self.num_sfb_tmem_cols,
                dtype=self.sf_dtype,
            )
            tCtSFB2 = cute.make_tensor(sfb2_tmem_ptr, tCtSFB_layout)
            (
                tiled_copy_s2t_sfa,
                tCsSFA_compact_s2t,
                tCtSFA_compact_s2t,
            ) = self.mainloop_s2t_copy_and_partition(sSFA, tCtSFA)
            (
                tiled_copy_s2t_sfb,
                tCsSFB1_compact_s2t,
                tCtSFB1_compact_s2t,
            ) = self.mainloop_s2t_copy_and_partition(sSFB1, tCtSFB1)
            (
                _,
                tCsSFB2_compact_s2t,
                tCtSFB2_compact_s2t,
            ) = self.mainloop_s2t_copy_and_partition(sSFB2, tCtSFB2)
            ab_consumer_state = pipeline.make_pipeline_state(
                pipeline.PipelineUserType.Consumer, self.num_ab_stage
            )
            acc_producer_state = pipeline.make_pipeline_state(
                pipeline.PipelineUserType.Producer, self.num_acc_stage
            )
            mma_tile_coord_mnl = (
                bidx // cute.size(tiled_mma.thr_id.shape),
                bidy,
                bidz,
            )
            if cutlass.const_expr(self.overlapping_accum):
                acc_stage_index = acc_producer_state.phase ^ 1
            else:
                acc_stage_index = acc_producer_state.index
            tCtAcc1 = tCtAcc1_base[(None, None, None, acc_stage_index)]
            tCtAcc2 = tCtAcc2_base[(None, None, None, acc_stage_index)]
            # Fused dual GEMM: both GEMMs per k-tile iteration
            total_k_tiles = k_tile_cnt
            ab_consumer_state.reset_count()
            peek_ab_full_status = cutlass.Boolean(1)
            if ab_consumer_state.count < total_k_tiles and is_leader_cta:
                peek_ab_full_status = ab_pipeline.consumer_try_wait(
                    ab_consumer_state
                )
            if is_leader_cta:
                acc_pipeline.producer_acquire(acc_producer_state)
            tCtSFB1_mma = tCtSFB1
            tCtSFB2_mma = tCtSFB2
            if cutlass.const_expr(self.cta_tile_shape_mnk[1] == 192):
                offset = cutlass.Int32(2) if mma_tile_coord_mnl[1] % 2 == 1 else cutlass.Int32(0)
                shifted_ptr1 = cute.recast_ptr(
                    acc_tmem_ptr
                     + self.num_accumulator_tmem_cols
                     + self.num_sfa_tmem_cols
                    + offset,
                    dtype=self.sf_dtype,
                )
                tCtSFB1_mma = cute.make_tensor(shifted_ptr1, tCtSFB_layout)
                shifted_ptr2 = cute.recast_ptr(
                    acc_tmem_ptr
                     + self.num_accumulator_tmem_cols
                     + self.num_sfa_tmem_cols
                     + self.num_sfb_tmem_cols
                    + offset,
                    dtype=self.sf_dtype,
                )
                tCtSFB2_mma = cute.make_tensor(shifted_ptr2, tCtSFB_layout)
            elif cutlass.const_expr(self.cta_tile_shape_mnk[1] == 64):
                offset = cutlass.Int32((mma_tile_coord_mnl[1] % 2) * 2)
                shifted_ptr1 = cute.recast_ptr(
                    acc_tmem_ptr
                    + self.num_accumulator_tmem_cols
                    + self.num_sfa_tmem_cols
                    + offset,
                    dtype=self.sf_dtype,
                )
                tCtSFB1_mma = cute.make_tensor(shifted_ptr1, tCtSFB_layout)
                shifted_ptr2 = cute.recast_ptr(
                    acc_tmem_ptr
                    + self.num_accumulator_tmem_cols
                    + self.num_sfa_tmem_cols
                    + self.num_sfb_tmem_cols
                    + offset,
                    dtype=self.sf_dtype,
                )
                tCtSFB2_mma = cute.make_tensor(shifted_ptr2, tCtSFB_layout)
            # Fused dual GEMM mainloop
            for k_tile in range(total_k_tiles):
                if is_leader_cta:
                    ab_pipeline.consumer_wait(
                        ab_consumer_state, peek_ab_full_status
                    )
                    s2t_stage_coord = (
                        None,
                        None,
                        None,
                        None,
                        ab_consumer_state.index,
                    )
                    tCsSFA_compact_s2t_staged = tCsSFA_compact_s2t[s2t_stage_coord]
                    tCsSFB1_compact_s2t_staged = tCsSFB1_compact_s2t[s2t_stage_coord]
                    tCsSFB2_compact_s2t_staged = tCsSFB2_compact_s2t[s2t_stage_coord]
                    cute.copy(
                        tiled_copy_s2t_sfa,
                        tCsSFA_compact_s2t_staged,
                        tCtSFA_compact_s2t,
                    )
                    # Copy both SFB1 and SFB2 to tmem
                    cute.copy(
                        tiled_copy_s2t_sfb,
                        tCsSFB1_compact_s2t_staged,
                        tCtSFB1_compact_s2t,
                    )
                    cute.copy(
                        tiled_copy_s2t_sfb,
                        tCsSFB2_compact_s2t_staged,
                        tCtSFB2_compact_s2t,
                    )
                    num_kblocks = cute.size(tCrA, mode=[2])
                    # GEMM1: A @ B1
                    tiled_mma.set(tcgen05.Field.ACCUMULATE, k_tile > 0)
                    for kblock_idx in cutlass.range(num_kblocks, unroll_full=True):
                        kblock_coord = (
                            None,
                            None,
                            kblock_idx,
                            ab_consumer_state.index,
                        )
                        sf_kblock_coord = (None, None, kblock_idx)
                        tiled_mma.set(
                            tcgen05.Field.SFA,
                            tCtSFA[sf_kblock_coord].iterator,
                        )
                        tiled_mma.set(
                            tcgen05.Field.SFB,
                            tCtSFB1_mma[sf_kblock_coord].iterator,
                        )
                        cute.gemm(
                            tiled_mma,
                            tCtAcc1,
                            tCrA[kblock_coord],
                            tCrB1[kblock_coord],
                            tCtAcc1,
                        )
                        tiled_mma.set(tcgen05.Field.ACCUMULATE, True)
                    # GEMM2: A @ B2
                    tiled_mma.set(tcgen05.Field.ACCUMULATE, k_tile > 0)
                    for kblock_idx in cutlass.range(num_kblocks, unroll_full=True):
                        kblock_coord = (
                            None,
                            None,
                            kblock_idx,
                            ab_consumer_state.index,
                        )
                        sf_kblock_coord = (None, None, kblock_idx)
                        tiled_mma.set(
                            tcgen05.Field.SFA,
                            tCtSFA[sf_kblock_coord].iterator,
                        )
                        tiled_mma.set(
                            tcgen05.Field.SFB,
                            tCtSFB2_mma[sf_kblock_coord].iterator,
                        )
                        cute.gemm(
                            tiled_mma,
                            tCtAcc2,
                            tCrA[kblock_coord],
                            tCrB2[kblock_coord],
                            tCtAcc2,
                        )
                        tiled_mma.set(tcgen05.Field.ACCUMULATE, True)
                    ab_pipeline.consumer_release(ab_consumer_state)
                ab_consumer_state.advance()
                peek_ab_full_status = cutlass.Boolean(1)
                if ab_consumer_state.count < total_k_tiles:
                    if is_leader_cta:
                        peek_ab_full_status = ab_pipeline.consumer_try_wait(
                            ab_consumer_state
                        )
            if is_leader_cta:
                acc_pipeline.producer_commit(acc_producer_state)
            acc_producer_state.advance()
            acc_pipeline.producer_tail(acc_producer_state)
        if warp_idx < self.mma_warp_id:
            tmem.allocate(self.num_tmem_alloc_cols)
            tmem.wait_for_alloc()
            acc_tmem_ptr = tmem.retrieve_ptr(self.acc_dtype)
            # Create two accumulator bases for dual GEMM
            tCtAcc1_base = cute.make_tensor(acc_tmem_ptr, tCtAcc_fake.layout)
            acc2_tmem_ptr = acc_tmem_ptr + self.num_accumulator_tmem_cols_single
            tCtAcc2_base = cute.make_tensor(acc2_tmem_ptr, tCtAcc_fake.layout)
            epi_tidx = tidx
            (
                tiled_copy_t2r,
                tTR_tAcc1_base,
                tTR_rAcc1,
            ) = self.epilog_tmem_copy_and_partition(
                epi_tidx, tCtAcc1_base, tCgC, epi_tile, use_2cta_instrs
            )
            # Second accumulator partition
            (
                _,
                tTR_tAcc2_base,
                tTR_rAcc2,
            ) = self.epilog_tmem_copy_and_partition(
                epi_tidx, tCtAcc2_base, tCgC, epi_tile, use_2cta_instrs
            )
            tTR_rC = cute.make_rmem_tensor(tTR_rAcc1.shape, self.c_dtype)
            tiled_copy_r2s, tRS_rC, tRS_sC = self.epilog_smem_copy_and_partition(
                tiled_copy_t2r, tTR_rC, epi_tidx, sC
            )
            (
                tma_atom_c,
                bSG_sC,
                bSG_gC_partitioned,
            ) = self.epilog_gmem_copy_and_partition(
                epi_tidx, tma_atom_c, tCgC, epi_tile, sC
            )
            acc_consumer_state = pipeline.make_pipeline_state(
                pipeline.PipelineUserType.Consumer, self.num_acc_stage
            )
            c_producer_group = pipeline.CooperativeGroup(
                pipeline.Agent.Thread,
                32 * len(self.epilog_warp_id),
            )
            c_pipeline = pipeline.PipelineTmaStore.create(
                num_stages=self.num_c_stage,
                producer_group=c_producer_group,
            )
            mma_tile_coord_mnl = (
                bidx // cute.size(tiled_mma.thr_id.shape),
                bidy,
                bidz,
            )
            bSG_gC = bSG_gC_partitioned[
                (
                    None,
                    None,
                    None,
                    *mma_tile_coord_mnl,
                )
            ]
            if cutlass.const_expr(self.overlapping_accum):
                acc_stage_index = acc_consumer_state.phase
                reverse_subtile = cutlass.Boolean(True) if acc_stage_index == 0 else cutlass.Boolean(False)
            else:
                acc_stage_index = acc_consumer_state.index
            tTR_tAcc1 = tTR_tAcc1_base[
                (None, None, None, None, None, acc_stage_index)
            ]
            tTR_tAcc2 = tTR_tAcc2_base[
                (None, None, None, None, None, acc_stage_index)
            ]
            acc_pipeline.consumer_wait(acc_consumer_state)
            tTR_tAcc1 = cute.group_modes(tTR_tAcc1, 3, cute.rank(tTR_tAcc1))
            tTR_tAcc2 = cute.group_modes(tTR_tAcc2, 3, cute.rank(tTR_tAcc2))
            bSG_gC = cute.group_modes(bSG_gC, 1, cute.rank(bSG_gC))
            subtile_cnt = cute.size(tTR_tAcc1.shape, mode=[3])
            num_prev_subtiles = cutlass.Int32(0)
            for subtile_idx in cutlass.range(subtile_cnt):
                real_subtile_idx = subtile_idx
                if cutlass.const_expr(self.overlapping_accum):
                    if reverse_subtile:
                        real_subtile_idx = self.cta_tile_shape_mnk[1] // self.epi_tile_n - 1 - subtile_idx
                tTR_tAcc1_mn = tTR_tAcc1[(None, None, None, real_subtile_idx)]
                tTR_tAcc2_mn = tTR_tAcc2[(None, None, None, real_subtile_idx)]
                # Load both accumulators
                cute.copy(tiled_copy_t2r, tTR_tAcc1_mn, tTR_rAcc1)
                cute.copy(tiled_copy_t2r, tTR_tAcc2_mn, tTR_rAcc2)
                if cutlass.const_expr(self.overlapping_accum):
                    if subtile_idx == self.iter_acc_early_release_in_epilogue:
                        cute.arch.fence_view_async_tmem_load()
                        with cute.arch.elect_one():
                            acc_pipeline.consumer_release(acc_consumer_state)
                        acc_consumer_state.advance()
                # Apply silu to first accumulator and multiply with second
                acc1_vec = tiled_copy_r2s.retile(tTR_rAcc1).load()
                acc2_vec = tiled_copy_r2s.retile(tTR_rAcc2).load()
                # silu(x) = x * sigmoid(x) where sigmoid(x) = 0.5 * (1 + tanh(x * 0.5))
                # silu(x) = 0.5 * x * (1 + tanh(x * 0.5))
                # This avoids expensive division
                half_x = acc1_vec * 0.5
                tanh_term = cute_math.tanh(half_x, fastmath=True)
                silu_result = half_x * (1.0 + tanh_term)
                result_vec = (silu_result * acc2_vec).to(self.c_dtype)
                tRS_rC.store(result_vec)
                c_buffer = (num_prev_subtiles + real_subtile_idx) % self.num_c_stage
                cute.copy(
                    tiled_copy_r2s,
                    tRS_rC,
                    tRS_sC[(None, None, None, c_buffer)],
                )
                cute.arch.fence_proxy(
                    cute.arch.ProxyKind.async_shared,
                    space=cute.arch.SharedSpace.shared_cta,
                )
                self.epilog_sync_barrier.arrive_and_wait()
                if warp_idx == self.epilog_warp_id[0]:
                    try:
                        cute.copy(
                            tma_atom_c,
                            bSG_sC[(None, c_buffer)],
                            bSG_gC[(None, real_subtile_idx)],
                            cache_policy=cutlass.Int64(cutlass.Int64(_TMA_CACHE_EVICT_FIRST).ir_value()),
                        )
                    except TypeError:
                        cute.copy(
                            tma_atom_c,
                            bSG_sC[(None, c_buffer)],
                            bSG_gC[(None, real_subtile_idx)],
                        )
                    c_pipeline.producer_commit()
                    c_pipeline.producer_acquire()
                self.epilog_sync_barrier.arrive_and_wait()
            if cutlass.const_expr(not self.overlapping_accum):
                with cute.arch.elect_one():
                    acc_pipeline.consumer_release(acc_consumer_state)
                acc_consumer_state.advance()
            tmem.relinquish_alloc_permit()
            self.epilog_sync_barrier.arrive_and_wait()
            tmem.free(acc_tmem_ptr)
            c_pipeline.producer_tail()
    def mainloop_s2t_copy_and_partition(
        self,
        sSF: cute.Tensor,
        tSF: cute.Tensor,
    ) -> Tuple[cute.TiledCopy, cute.Tensor, cute.Tensor]:
        tCsSF_compact = cute.filter_zeros(sSF)
        tCtSF_compact = cute.filter_zeros(tSF)
        copy_atom_s2t = cute.make_copy_atom(
            tcgen05.Cp4x32x128bOp(self.cta_group),
            self.sf_dtype,
        )
        tiled_copy_s2t = tcgen05.make_s2t_copy(copy_atom_s2t, tCtSF_compact)
        thr_copy_s2t = tiled_copy_s2t.get_slice(0)
        tCsSF_compact_s2t_ = thr_copy_s2t.partition_S(tCsSF_compact)
        tCsSF_compact_s2t = tcgen05.get_s2t_smem_desc_tensor(
            tiled_copy_s2t, tCsSF_compact_s2t_
        )
        tCtSF_compact_s2t = thr_copy_s2t.partition_D(tCtSF_compact)
        return tiled_copy_s2t, tCsSF_compact_s2t, tCtSF_compact_s2t
    def epilog_tmem_copy_and_partition(
        self,
        tidx: cutlass.Int32,
        tAcc: cute.Tensor,
        gC_mnl: cute.Tensor,
        epi_tile: cute.Tile,
        use_2cta_instrs: Union[cutlass.Boolean, bool],
    ) -> Tuple[cute.TiledCopy, cute.Tensor, cute.Tensor]:
        copy_atom_t2r = sm100_utils.get_tmem_load_op(
            self.cta_tile_shape_mnk,
            self.c_layout,
            self.c_dtype,
            self.acc_dtype,
            epi_tile,
            use_2cta_instrs,
        )
        tAcc_epi = cute.flat_divide(
            tAcc[((None, None), 0, 0, None)],
            epi_tile,
        )
        tiled_copy_t2r = tcgen05.make_tmem_copy(
            copy_atom_t2r, tAcc_epi[(None, None, 0, 0, 0)]
        )
        thr_copy_t2r = tiled_copy_t2r.get_slice(tidx)
        tTR_tAcc = thr_copy_t2r.partition_S(tAcc_epi)
        gC_mnl_epi = cute.flat_divide(
            gC_mnl[((None, None), 0, 0, None, None, None)], epi_tile
        )
        tTR_gC = thr_copy_t2r.partition_D(gC_mnl_epi)
        tTR_rAcc = cute.make_rmem_tensor(
            tTR_gC[(None, None, None, 0, 0, 0, 0, 0)].shape, self.acc_dtype
        )
        return tiled_copy_t2r, tTR_tAcc, tTR_rAcc
    def epilog_smem_copy_and_partition(
        self,
        tiled_copy_t2r: cute.TiledCopy,
        tTR_rC: cute.Tensor,
        tidx: cutlass.Int32,
        sC: cute.Tensor,
    ) -> Tuple[cute.TiledCopy, cute.Tensor, cute.Tensor]:
        copy_atom_r2s = sm100_utils.get_smem_store_op(
            self.c_layout, self.c_dtype, self.acc_dtype, tiled_copy_t2r
        )
        tiled_copy_r2s = cute.make_tiled_copy_D(copy_atom_r2s, tiled_copy_t2r)
        thr_copy_r2s = tiled_copy_r2s.get_slice(tidx)
        tRS_sC = thr_copy_r2s.partition_D(sC)
        tRS_rC = tiled_copy_r2s.retile(tTR_rC)
        return tiled_copy_r2s, tRS_rC, tRS_sC
    def epilog_gmem_copy_and_partition(
        self,
        tidx: cutlass.Int32,
        atom: Union[cute.CopyAtom, cute.TiledCopy],
        gC_mnl: cute.Tensor,
        epi_tile: cute.Tile,
        sC: cute.Tensor,
    ) -> Tuple[cute.CopyAtom, cute.Tensor, cute.Tensor]:
        gC_epi = cute.flat_divide(
            gC_mnl[((None, None), 0, 0, None, None, None)], epi_tile
        )
        tma_atom_c = atom
        sC_for_tma_partition = cute.group_modes(sC, 0, 2)
        gC_for_tma_partition = cute.group_modes(gC_epi, 0, 2)
        bSG_sC, bSG_gC = cpasync.tma_partition(
            tma_atom_c,
            0,
            cute.make_layout(1),
            sC_for_tma_partition,
            gC_for_tma_partition,
        )
        return tma_atom_c, bSG_sC, bSG_gC
    @staticmethod
    def _compute_stages(
        tiled_mma: cute.TiledMma,
        mma_tiler_mnk: Tuple[int, int, int],
        a_dtype: Type[cutlass.Numeric],
        b_dtype: Type[cutlass.Numeric],
        epi_tile: cute.Tile,
        c_dtype: Type[cutlass.Numeric],
        c_layout: utils.LayoutEnum,
        sf_dtype: Type[cutlass.Numeric],
        sf_vec_size: int,
        smem_capacity: int,
        occupancy: int,
    ) -> Tuple[int, int, int]:
        num_acc_stage = 1 if mma_tiler_mnk[1] == 256 else 2
        num_c_stage = 2
        a_smem_layout_stage_one = sm100_utils.make_smem_layout_a(
            tiled_mma,
            mma_tiler_mnk,
            a_dtype,
            1,  # a tmp 1 stage is provided
        )
        b_smem_layout_staged_one = sm100_utils.make_smem_layout_b(
            tiled_mma,
            mma_tiler_mnk,
            b_dtype,
            1,  # a tmp 1 stage is provided
        )
        sfa_smem_layout_staged_one = blockscaled_utils.make_smem_layout_sfa(
            tiled_mma,
            mma_tiler_mnk,
            sf_vec_size,
            1,  # a tmp 1 stage is provided
        )
        sfb_smem_layout_staged_one = blockscaled_utils.make_smem_layout_sfb(
            tiled_mma,
            mma_tiler_mnk,
            sf_vec_size,
            1,  # a tmp 1 stage is provided
        )
        c_smem_layout_staged_one = sm100_utils.make_smem_layout_epi(
            c_dtype,
            c_layout,
            epi_tile,
            1,
        )
        # Fused dual GEMM: need double B and SFB buffers
        ab_bytes_per_stage = (
            cute.size_in_bytes(a_dtype, a_smem_layout_stage_one)
            + cute.size_in_bytes(b_dtype, b_smem_layout_staged_one) * 2  # B1 and B2
            + cute.size_in_bytes(sf_dtype, sfa_smem_layout_staged_one)
            + cute.size_in_bytes(sf_dtype, sfb_smem_layout_staged_one) * 2  # SFB1 and SFB2
        )
        mbar_helpers_bytes = 1024
        c_bytes_per_stage = cute.size_in_bytes(c_dtype, c_smem_layout_staged_one)
        c_bytes = c_bytes_per_stage * num_c_stage
        num_ab_stage = (
            smem_capacity // occupancy - (mbar_helpers_bytes + c_bytes)
        ) // ab_bytes_per_stage
        num_c_stage += (
            smem_capacity
            - occupancy * ab_bytes_per_stage * num_ab_stage
            - occupancy * (mbar_helpers_bytes + c_bytes)
        ) // (occupancy * c_bytes_per_stage)
        return num_acc_stage, num_ab_stage, num_c_stage
    @staticmethod
    def _compute_grid(
        c: cute.Tensor,
        cta_tile_shape_mnk: Tuple[int, int, int],
        cluster_shape_mn: Tuple[int, int],
        max_active_clusters: cutlass.Constexpr,
    ) -> Tuple[utils.PersistentTileSchedulerParams, Tuple[int, int, int]]:
        c_shape = cute.slice_(cta_tile_shape_mnk, (None, None, 0))
        gc = cute.zipped_divide(c, tiler=c_shape)
        num_ctas_mnl = gc[(0, (None, None, None))].shape
        cluster_shape_mnl = (*cluster_shape_mn, 1)
        tile_sched_params = utils.PersistentTileSchedulerParams(
            num_ctas_mnl, cluster_shape_mnl
        )
        grid = utils.StaticPersistentTileScheduler.get_grid_shape(
            tile_sched_params, max_active_clusters
        )
        return tile_sched_params, grid
    @staticmethod
    def is_valid_dtypes_and_scale_factor_vec_size(
        ab_dtype: Type[cutlass.Numeric],
        sf_dtype: Type[cutlass.Numeric],
        sf_vec_size: int,
        c_dtype: Type[cutlass.Numeric],
    ) -> bool:
        is_valid = True
        if ab_dtype not in {
            cutlass.Float4E2M1FN,
            cutlass.Float8E5M2,
            cutlass.Float8E4M3FN,
        }:
            is_valid = False
        if sf_vec_size not in {16, 32}:
            is_valid = False
        if sf_dtype not in {cutlass.Float8E8M0FNU, cutlass.Float8E4M3FN}:
            is_valid = False
        if sf_dtype == cutlass.Float8E4M3FN and sf_vec_size == 32:
            is_valid = False
        if ab_dtype in {cutlass.Float8E5M2, cutlass.Float8E4M3FN} and sf_vec_size == 16:
            is_valid = False
        if c_dtype not in {
            cutlass.Float32,
            cutlass.Float16,
            cutlass.BFloat16,
            cutlass.Float8E5M2,
            cutlass.Float8E4M3FN,
        }:
            is_valid = False
        return is_valid
    @staticmethod
    def is_valid_layouts(
        ab_dtype: Type[cutlass.Numeric],
        c_dtype: Type[cutlass.Numeric],
        a_major: str,
        b_major: str,
        c_major: str,
    ) -> bool:
        is_valid = True
        if ab_dtype is cutlass.Float4E2M1FN and not (a_major == "k" and b_major == "k"):
            is_valid = False
        return is_valid
    @staticmethod
    def is_valid_mma_tiler_and_cluster_shape(
        mma_tiler_mn: Tuple[int, int],
        cluster_shape_mn: Tuple[int, int],
    ) -> bool:
        is_valid = True
        if mma_tiler_mn[0] not in [128, 256]:
            is_valid = False
        if mma_tiler_mn[1] not in [64, 128, 192, 256]:
            is_valid = False
        if cluster_shape_mn[0] % (2 if mma_tiler_mn[0] == 256 else 1) != 0:
            is_valid = False
        is_power_of_2 = lambda x: x > 0 and (x & (x - 1)) == 0
        if (
            cluster_shape_mn[0] * cluster_shape_mn[1] > 16
            or cluster_shape_mn[0] <= 0
            or cluster_shape_mn[1] <= 0
            or cluster_shape_mn[0] > 4
            or cluster_shape_mn[1] > 4
            or not is_power_of_2(cluster_shape_mn[0])
            or not is_power_of_2(cluster_shape_mn[1])
        ):
            is_valid = False
        return is_valid
    @staticmethod
    def is_valid_tensor_alignment(
        m: int,
        n: int,
        k: int,
        l: int,
        ab_dtype: Type[cutlass.Numeric],
        c_dtype: Type[cutlass.Numeric],
        a_major: str,
        b_major: str,
        c_major: str,
    ) -> bool:
        is_valid = True
        def check_contigous_16B_alignment(dtype, is_mode0_major, tensor_shape):
            major_mode_idx = 0 if is_mode0_major else 1
            num_major_elements = tensor_shape[major_mode_idx]
            num_contiguous_elements = 16 * 8 // dtype.width
            return num_major_elements % num_contiguous_elements == 0
        if (
            not check_contigous_16B_alignment(ab_dtype, a_major == "m", (m, k, l))
            or not check_contigous_16B_alignment(ab_dtype, b_major == "n", (n, k, l))
            or not check_contigous_16B_alignment(c_dtype, c_major == "m", (m, n, l))
        ):
            is_valid = False
        return is_valid
    @staticmethod
    def can_implement(
        ab_dtype: Type[cutlass.Numeric],
        sf_dtype: Type[cutlass.Numeric],
        sf_vec_size: int,
        c_dtype: Type[cutlass.Numeric],
        mma_tiler_mn: Tuple[int, int],
        cluster_shape_mn: Tuple[int, int],
        m: int,
        n: int,
        k: int,
        l: int,
        a_major: str,
        b_major: str,
        c_major: str,
    ) -> bool:
        can_implement = True
        if not Sm100BlockScaledPersistentDenseGemmKernel.is_valid_dtypes_and_scale_factor_vec_size(
            ab_dtype, sf_dtype, sf_vec_size, c_dtype
        ):
            can_implement = False
        if not Sm100BlockScaledPersistentDenseGemmKernel.is_valid_layouts(
            ab_dtype, c_dtype, a_major, b_major, c_major
        ):
            can_implement = False
        if not Sm100BlockScaledPersistentDenseGemmKernel.is_valid_mma_tiler_and_cluster_shape(
            mma_tiler_mn, cluster_shape_mn
        ):
            can_implement = False
        if not Sm100BlockScaledPersistentDenseGemmKernel.is_valid_tensor_alignment(
            m, n, k, l, ab_dtype, c_dtype, a_major, b_major, c_major
        ):
            can_implement = False
        return can_implement
_compiled = None
_max_active_clusters = None
def _compile_kernel():
    global _compiled, _max_active_clusters
    if _compiled is not None:
        return _compiled
    _max_active_clusters = 4096
    gemm = Sm100BlockScaledPersistentDenseGemmKernel(
        _SF_VEC_SIZE, _MMA_TILER_MN, _CLUSTER_SHAPE_MN, occupancy=_OCCUPANCY
    )
    a_ptr = make_ptr(cutlass.Float4E2M1FN, 0, cute.AddressSpace.gmem, assumed_align=128)
    b1_ptr = make_ptr(cutlass.Float4E2M1FN, 0, cute.AddressSpace.gmem, assumed_align=128)
    b2_ptr = make_ptr(cutlass.Float4E2M1FN, 0, cute.AddressSpace.gmem, assumed_align=128)
    sfa_ptr = make_ptr(cutlass.Float8E4M3FN, 0, cute.AddressSpace.gmem, assumed_align=128)
    sfb1_ptr = make_ptr(cutlass.Float8E4M3FN, 0, cute.AddressSpace.gmem, assumed_align=128)
    sfb2_ptr = make_ptr(cutlass.Float8E4M3FN, 0, cute.AddressSpace.gmem, assumed_align=128)
    c_ptr = make_ptr(cutlass.Float16, 0, cute.AddressSpace.gmem, assumed_align=128)
    _compiled = cute.compile(
        gemm,
        a_ptr, b1_ptr, b2_ptr, sfa_ptr, sfb1_ptr, sfb2_ptr, c_ptr,
        (0, 0, 0, 0),
        _max_active_clusters,
        options="--opt-level 3",
    )
    return _compiled

def custom_kernel(data):
    # Dual GEMM input: (a, b1, b2, sfa_cpu, sfb1_cpu, sfb2_cpu, sfa_p, sfb1_p, sfb2_p, c)
    a, b1, b2, _sfa, _sfb1, _sfb2, sfa_p, sfb1_p, sfb2_p, c = data
    compiled = _compile_kernel()
    m, k_half, l = a.shape
    n, k_half_b1, l_b1 = b1.shape
    n2, k_half_b2, l_b2 = b2.shape
    if int(l_b1) != int(l) or int(l_b2) != int(l):
        raise ValueError("a/b1/b2 L mismatch")
    k = int(k_half) * 2
    if int(k_half_b1) * 2 != k or int(k_half_b2) * 2 != k:
        raise ValueError("a/b1/b2 K mismatch")
    if n != n2:
        raise ValueError("b1/b2 N mismatch")
    a_ptr = make_ptr(cutlass.Float4E2M1FN, a.data_ptr(), cute.AddressSpace.gmem, assumed_align=128)
    b1_ptr = make_ptr(cutlass.Float4E2M1FN, b1.data_ptr(), cute.AddressSpace.gmem, assumed_align=128)
    b2_ptr = make_ptr(cutlass.Float4E2M1FN, b2.data_ptr(), cute.AddressSpace.gmem, assumed_align=128)
    sfa_ptr = make_ptr(cutlass.Float8E4M3FN, sfa_p.data_ptr(), cute.AddressSpace.gmem, assumed_align=128)
    sfb1_ptr = make_ptr(cutlass.Float8E4M3FN, sfb1_p.data_ptr(), cute.AddressSpace.gmem, assumed_align=128)
    sfb2_ptr = make_ptr(cutlass.Float8E4M3FN, sfb2_p.data_ptr(), cute.AddressSpace.gmem, assumed_align=128)
    c_ptr = make_ptr(cutlass.Float16, c.data_ptr(), cute.AddressSpace.gmem, assumed_align=128)
    compiled(a_ptr, b1_ptr, b2_ptr, sfa_ptr, sfb1_ptr, sfb2_ptr, c_ptr, (m, n, k, l))
    return c

def compile_kernel():
    """Public function to compile the kernel, called by eval.py"""
    return _compile_kernel()