For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo function

sm100_heuristic_and_outliers_dispatch

def sm100_heuristic_and_outliers_dispatch[c_type: DType, a_type: DType, b_type: DType, //, transpose_b: Bool = True, elementwise_lambda_fn: Optional[def[dtype: DType, width: SIMDLength, *, alignment: Int = Int(1)](IndexList[Int(2)], SIMD[dtype, width]) capturing thin -> None] = None, elementwise_compute_lambda_fn: Optional[def[dtype: DType, width: SIMDLength, *, alignment: Int = Int(1)](IndexList[Int(2)], SIMD[dtype, width]) capturing thin -> SIMD[dtype, width]] = None, pdl_level: PDLLevel = PDLLevel(), has_epilogue_tensor: Bool = False, epilogue_is_1d: Bool = False](c: TileTensor[c_type, Storage=c.Storage, address_space=c.address_space, linear_idx_type=c.linear_idx_type], a: TileTensor[a_type, Storage=a.Storage, address_space=a.address_space, linear_idx_type=a.linear_idx_type], b: TileTensor[b_type, Storage=b.Storage, address_space=b.address_space, linear_idx_type=b.linear_idx_type], ctx: DeviceContext, epilogue_tensor: OptionalReg[TileTensor[c_type, Layout[*?, *?], ImmutAnyOrigin]] = None) -> Int

Dispatches an SM100 matmul through the heuristic outlier config set.

Wraps select_and_launch_sm100_config with a launch callback that invokes blackwell_matmul_tma_umaa_warp_specialized directly, passing through the elementwise and compute epilogue lambdas.

Parameters:

c_type (DType): Output element type (inferred).
a_type (DType): Element type of the LHS operand a (inferred).
b_type (DType): Element type of the RHS operand b (inferred).
transpose_b (Bool): Whether b is stored transposed (defaults to True).
elementwise_lambda_fn (Optional[def[dtype: DType, width: SIMDLength, *, alignment: Int = Int(1)](IndexList[Int(2)], SIMD[dtype, width]) capturing thin -> None]): Optional epilogue applied to each output element (defaults to None).
elementwise_compute_lambda_fn (Optional[def[dtype: DType, width: SIMDLength, *, alignment: Int = Int(1)](IndexList[Int(2)], SIMD[dtype, width]) capturing thin -> SIMD[dtype, width]]): Optional compute epilogue lambda, for example a static scale (defaults to None).
pdl_level (PDLLevel): Programmatic dependent launch level for the dispatched kernel (defaults to PDLLevel()).
has_epilogue_tensor (Bool): Whether an epilogue tensor is supplied for the TMA epilogue load path (defaults to False).
epilogue_is_1d (Bool): Whether the epilogue tensor is treated as 1D rather than row-major 2D (defaults to False).

Returns:

Int