Mojo function
dispatch_amd_matmul_by_block_shape
dispatch_amd_matmul_by_block_shape[c_type: DType, a_type: DType, b_type: DType, transpose_b: Bool, N: Int, K: Int, launcher_fn: fn[config: MatmulConfig[a_type, b_type, c_type, transpose_b]]() raises capturing -> None, default_block_tile_shape: IndexList[3]](M: Int, ctx: DeviceContext)
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!