IMPORTANT: To view this page as Markdown, append `.md` to the URL (e.g. /get-started.md). For the complete documentation index, see llms.txt.
Skip to main content
For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Mojo function

matmul_sm100_fallback

def matmul_sm100_fallback[ComputeFnType: ElementwiseComputeFn = inflated$def[comptime[dtype: DType, width: SIMDLength, *, alignment: Int] #kgen.func.symbol<@linalg::@utils::@"identity_compute_fn[::DType,::SIMDLength,::SIMD[DType.int, 1],KGENParamList[::Intable & ::Defaultable & ::RegisterPassable & ::TrivialRegisterPassable & ::Writable & ::AnyType & ::Copyable & ::ImplicitlyCopyable & ::Deinitable & ::Movable & ::CoordLike],::TypeList[::Intable & ::Defaultable & ::RegisterPassable & ::TrivialRegisterPassable & ::Writable & ::AnyType & ::Copyable & ::ImplicitlyCopyable & ::Deinitable & ::Movable & ::CoordLike, $3]](::Coord[$3, $4],::SIMD[$0, $1])"<:!lit.struct<@std::@builtin::@dtype::@DType> dtype, :!lit.struct<@std::@builtin::@simd_length::@SIMDLength> width, :!lit.struct<@std::@simd::@SIMD<:!lit.struct<@std::@builtin::@dtype::@DType> {:dtype index}, :!lit.struct<@std::@builtin::@simd_length::@SIMDLength> {1}>> alignment, :param_list<trait<@std::@builtin::@int::@Intable, @std::@builtin::@value::@Defaultable, @std::@builtin::@value::@RegisterPassable, @std::@builtin::@value::@TrivialRegisterPassable, @std::@format::@__init__::@Writable, @std::@traits::@anytype::@AnyType, @std::@traits::@copyable::@Copyable, @std::@traits::@copyable::@ImplicitlyCopyable, @std::@traits::@deinitable::@Deinitable, @std::@traits::@movable::@Movable, @std::@utils::@coord::@CoordLike>> *?, :!lit.struct<@std::@builtin::@variadics::@TypeList<:meta<!lit.trait<@std::@traits::@anytype::@AnyType>> trait<@std::@builtin::@int::@Intable, @std::@builtin::@value::@Defaultable, @std::@builtin::@value::@RegisterPassable, @std::@builtin::@value::@TrivialRegisterPassable, @std::@format::@__init__::@Writable, @std::@traits::@anytype::@AnyType, @std::@traits::@copyable::@Copyable, @std::@traits::@copyable::@ImplicitlyCopyable, @std::@traits::@deinitable::@Deinitable, @std::@traits::@movable::@Movable, @std::@utils::@coord::@CoordLike>, :param_list<trait<@std::@builtin::@int::@Intable, @std::@builtin::@value::@Defaultable, @std::@builtin::@value::@RegisterPassable, @std::@builtin::@value::@TrivialRegisterPassable, @std::@format::@__init__::@Writable, @std::@traits::@anytype::@AnyType, @std::@traits::@copyable::@Copyable, @std::@traits::@copyable::@ImplicitlyCopyable, @std::@traits::@deinitable::@Deinitable, @std::@traits::@movable::@Movable, @std::@utils::@coord::@CoordLike>> *?>> *?>>], //, c_type: DType, a_type: DType, b_type: DType, *, transpose_b: Bool, umma_shape: Tuple[Int, Int, Int], block_tile_shape: Tuple[Int, Int, Int], a_swizzle: TensorMapSwizzle = TensorMapSwizzle.SWIZZLE_128B, b_swizzle: TensorMapSwizzle = TensorMapSwizzle.SWIZZLE_128B, has_compute_fn: Bool = False](c: TileTensor[Engine=c.Engine, address_space=c.address_space, linear_idx_type=c.linear_idx_type], a: TileTensor[Engine=a.Engine, address_space=a.address_space, linear_idx_type=a.linear_idx_type], b: TileTensor[Engine=b.Engine, address_space=b.address_space, linear_idx_type=b.linear_idx_type], ctx: DeviceContext, compute_fn: ComputeFnType = no_compute_fn)

Launches the SM100 fallback matmul kernel for unsupported shapes or dtypes.

Uses a simple non-pipelined kernel without CLC/TMEM overhead, computing shared memory from the actual tile sizes rather than the hardware maximum. Only transposed-B, bfloat16, and float8_e4m3fn inputs are supported.

Parameters:

  • ​ComputeFnType (ElementwiseComputeFn): Type of compute_fn (inferred).
  • ​c_type (DType): Output element dtype of the result matrix.
  • ​a_type (DType): Input element dtype of the LHS matrix; must be bfloat16 or float8_e4m3fn.
  • ​b_type (DType): Input element dtype of the RHS matrix; must equal a_type.
  • ​transpose_b (Bool): Whether B is stored transposed as (N, K). Must be True.
  • ​umma_shape (Tuple[Int, Int, Int]): Tensor core MMA instruction shape as a 3-element IndexList of (M, N, K).
  • ​block_tile_shape (Tuple[Int, Int, Int]): CTA block tile shape as a 3-element IndexList of (BM, BN, BK) giving the tile dimensions along M, N, and K.
  • ​a_swizzle (TensorMapSwizzle): TMA swizzle mode for A tensor loads (defaults to SWIZZLE_128B).
  • ​b_swizzle (TensorMapSwizzle): TMA swizzle mode for B tensor loads (defaults to SWIZZLE_128B).
  • ​has_compute_fn (Bool): Whether to apply compute_fn to each output value before it is stored (defaults to False).

Args:

Was this page helpful?