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

dispatch_amd_matmul_by_block_shape

def dispatch_amd_matmul_by_block_shape[c_type: DType, a_type: DType, b_type: DType, transpose_b: Bool, N: Int, K: Int, default_block_tile_shape: IndexList[3], use_heuristic: Bool = False, *, LauncherFnType: ImplicitlyCopyable & def[config: MatmulConfig[a_type, b_type, c_type, transpose_b]]() raises -> None](launcher_fn: LauncherFnType, M: Int, ctx: DeviceContext) where (eq LauncherFnType.a_type, a_type) where (eq LauncherFnType.b_type, b_type) where (eq LauncherFnType.c_type, c_type) where (eq LauncherFnType.transpose_b, transpose_b)

Dispatches to the best kernel configuration based on runtime M dimension.

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit