LayoutTensor Version

Overview

Implement a kernel that adds 10 to each position of 2D LayoutTensor a and stores it in 2D LayoutTensor out.

Note: You have more threads than positions.

Key concepts

In this puzzle, you’ll learn about:

• Using LayoutTensor for 2D array access
• Direct 2D indexing with tensor[i, j]
• Handling bounds checking with LayoutTensor

The key insight is that LayoutTensor provides a natural 2D indexing interface, abstracting away the underlying memory layout while still requiring bounds checking.

• 2D access: Natural \((i,j)\) indexing with LayoutTensor
• Memory abstraction: No manual row-major calculation needed
• Guard condition: Still need bounds checking in both dimensions
• Thread bounds: More threads \((3 \times 3)\) than tensor elements \((2 \times 2)\)

Code to complete

alias SIZE = 2
alias BLOCKS_PER_GRID = 1
alias THREADS_PER_BLOCK = (3, 3)
alias dtype = DType.float32
alias layout = Layout.row_major(SIZE, SIZE)


fn add_10_2d(
    out: LayoutTensor[mut=True, dtype, layout],
    a: LayoutTensor[mut=True, dtype, layout],
    size: Int,
):
    local_i = thread_idx.x
    local_j = thread_idx.y
    # FILL ME IN (roughly 2 lines)

View full file: problems/p04/p04_layout_tensor.mojo

Running the code

To test your solution, run the following command in your terminal:

magic run p04_layout_tensor

Your output will look like this if the puzzle isn’t solved yet:

out: HostBuffer([0.0, 0.0, 0.0, 0.0])
expected: HostBuffer([10.0, 11.0, 12.0, 13.0])

Solution

fn add_10_2d(
    out: LayoutTensor[mut=True, dtype, layout],
    a: LayoutTensor[mut=True, dtype, layout],
    size: Int,
):
    local_i = thread_idx.x
    local_j = thread_idx.y
    if local_i < size and local_j < size:
        out[local_i, local_j] = a[local_i, local_j] + 10.0