For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Mojo struct

MatmulSchedule

struct MatmulSchedule

Tile scheduling strategy for GPU matmul kernels.

Selects how thread blocks are mapped to output tiles across the grid. TILE1D uses a simple linear mapping; TILE2D adds wave-level 2D swizzling for better L2 locality; DS_SCHEDULER uses the DeepSeek persistent block scheduler; NONE disables the scheduler (non-persistent kernels).

Implemented traits

AnyType, Copyable, Deinitable, ImplicitlyCopyable, Movable, RegisterPassable, TrivialRegisterPassable

comptime members

DS_SCHEDULER

comptime DS_SCHEDULER = MatmulSchedule(Int32(3))

NONE

comptime NONE = MatmulSchedule(Int32(0))

TILE1D

comptime TILE1D = MatmulSchedule(Int32(1))

TILE2D

comptime TILE2D = MatmulSchedule(Int32(2))

Methods

__eq__

def __eq__(self, other: Self) -> Bool

Returns:

Bool

__ne__

def __ne__(self, other: Self) -> Bool

Returns:

Bool