Mojo function

enqueue_apple_matmul

enqueue_apple_matmul[in_type: DType, transpose_b: Bool = False](c: TileTensor[DType.float32, address_space=c.address_space, linear_idx_type=c.linear_idx_type, element_size=c.element_size], a: TileTensor[in_type, address_space=a.address_space, linear_idx_type=a.linear_idx_type, element_size=a.element_size], b: TileTensor[in_type, address_space=b.address_space, linear_idx_type=b.linear_idx_type, element_size=b.element_size], ctx: DeviceContext)

Enqueue the Apple M5 matmul kernel on the given device context.

Accepts row-major TileTensor operands. For transpose_b=True, B is expected with shape (N, K).

Raises:

If the attached GPU is not Apple M5 (compute_capability != 5). M1-M4 lack GPU neural accelerator; future generations require re-validation.

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