diff --git a/examples/python/CuTeDSL/distributed/distributed_all_gather_gemm_blackwell.py b/examples/python/CuTeDSL/distributed/distributed_all_gather_gemm_blackwell.py index 8ea68dfd..df077f56 100644 --- a/examples/python/CuTeDSL/distributed/distributed_all_gather_gemm_blackwell.py +++ b/examples/python/CuTeDSL/distributed/distributed_all_gather_gemm_blackwell.py @@ -1151,12 +1151,6 @@ class PersistentDenseGemmKernel: tile_sched_params, cute.arch.block_idx(), cute.arch.grid_dim() ) if cutlass.const_expr(self.use_tma_store): - # {$nv-internal-release begin} - # (EPI_TILE_M, EPI_TILE_N, STAGE) - # sC = storage.sC.get_tensor( - # c_smem_layout_staged.outer, swizzle=c_smem_layout_staged.inner - # ) - # {$nv-internal-release end} assert tma_atom_c is not None and sC is not None self.epilogue_tma_store( diff --git a/examples/python/CuTeDSL/distributed/distributed_gemm_all_reduce_blackwell.py b/examples/python/CuTeDSL/distributed/distributed_gemm_all_reduce_blackwell.py index 4ff104d6..8e1b00d1 100644 --- a/examples/python/CuTeDSL/distributed/distributed_gemm_all_reduce_blackwell.py +++ b/examples/python/CuTeDSL/distributed/distributed_gemm_all_reduce_blackwell.py @@ -1056,12 +1056,6 @@ class PersistentDenseGemmKernel: ) if cutlass.const_expr(self.use_tma_store): - # {$nv-internal-release begin} - # (EPI_TILE_M, EPI_TILE_N, STAGE) - # sC = storage.sC.get_tensor( - # c_smem_layout_staged.outer, swizzle=c_smem_layout_staged.inner - # ) - # {$nv-internal-release end} assert tma_atom_c is not None and sC is not None self.epilogue_tma_store_release_flag(