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

parallelize_over_rows

def parallelize_over_rows[func: def(Int, Int) capturing thin -> None](shape: IndexList[element_type=shape.element_type], axis: Int, grain_size: Int, ctx: Optional[DeviceContext] = None)

Parallelize func over non-axis dims of shape.

Parameters:

func (def(Int, Int) capturing thin -> None): Function to call on range of rows.

Args:

shape (IndexList[element_type=shape.element_type]): Shape to parallelize over.
axis (Int): Rows are slices along the axis dimension of shape.
grain_size (Int): The minimum number of elements to warrant using an additional thread.
ctx (Optional[DeviceContext]): Optional CPU DeviceContext to execute the work on.

def parallelize_over_rows[FuncType: def(Int, Int) -> None](func: FuncType, shape: IndexList[element_type=shape.element_type], axis: Int, grain_size: Int, ctx: Optional[DeviceContext] = None)

Parallelize func over non-axis dims of shape.

Parameters:

FuncType (def(Int, Int) -> None): The body function type.

Args:

func (FuncType): The closure carrying the captured state of the body function.
shape (IndexList[element_type=shape.element_type]): Shape to parallelize over.
axis (Int): Rows are slices along the axis dimension of shape.
grain_size (Int): The minimum number of elements to warrant using an additional thread.
ctx (Optional[DeviceContext]): Optional CPU DeviceContext to execute the work on.