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

fused_silu_kernel

def fused_silu_kernel[output_dtype: DType, input_dtype: DType, output_layout: TensorLayout, input_layout: TensorLayout, row_offsets_layout: TensorLayout, num_threads: Int, num_sms: Int](output_tensor: TileTensor[output_dtype, output_layout, MutUntrackedOrigin], input_tensor: TileTensor[input_dtype, input_layout, ImmUntrackedOrigin], row_offsets: TileTensor[DType.uint32, row_offsets_layout, ImmUntrackedOrigin])

This kernel performs the SILU operation for all the MLPs in the EP MoE module. We need to manually implement the kernel here is because after the EP dispatch phase, the actual number of received tokens is not known to the host. This kernel will read the row offsets to determine the actual number of received tokens in the input tensor.

Arguments: output_tensor: The output tensor to store the result. input_tensor: The input tensor to perform the SILU operation. row_offsets: The row offsets to determine the actual number of received tokens.

Parameters:

output_dtype (DType): Element type of the output_tensor (e.g. DType.bfloat16).
input_dtype (DType): Element type of the input_tensor; its accumulation type must be floating-point.
output_layout (TensorLayout): Layout of the output_tensor TileTensor.
input_layout (TensorLayout): Layout of the input_tensor TileTensor.
row_offsets_layout (TensorLayout): Layout of the 1D row_offsets TileTensor.
num_threads (Int): Number of threads per block; sets the MAX_THREADS_PER_BLOCK launch metadata.
num_sms (Int): Number of streaming multiprocessors (SMs) used to scatter processing across thread blocks.