Skip to main content

/

Docs

Nightly

v25.1

Log in

Python module

driver

Exposes APIs for interacting with hardware, such as allocating tensors on a GPU and moving tensors between the CPU and GPU. It provides interfaces for memory management, device properties, and hardware monitoring. Through these APIs, you can control data placement, track resource utilization, and configure device settings for optimal performance.

Accelerator()​

max.driver.Accelerator(id: int = -1) → Device

Creates an accelerator device with the specified ID.

Provides access to GPU or other hardware accelerators in the system.

from max import driver

device = driver.Accelerator()

# Or specify GPU id
device = driver.Accelerator(id=0)  # First GPU
device = driver.Accelerator(id=1)  # Second GPU

# Get device id
device_id = device.id
from max import driver

device = driver.Accelerator()

# Or specify GPU id
device = driver.Accelerator(id=0)  # First GPU
device = driver.Accelerator(id=1)  # Second GPU

# Get device id
device_id = device.id

CPU()​

max.driver.CPU(id: int = -1) → Device

Creates a CPU device for the specified NUMA node.

from max import driver

# Create default CPU device
device = driver.CPU()

# Or specify NUMA node id if using NUMA architecture
device = driver.CPU(id=0)  # First NUMA node
device = driver.CPU(id=1)  # Second NUMA node

# Get device id
device_id = device.id
from max import driver

# Create default CPU device
device = driver.CPU()

# Or specify NUMA node id if using NUMA architecture
device = driver.CPU(id=0)  # First NUMA node
device = driver.CPU(id=1)  # Second NUMA node

# Get device id
device_id = device.id

DLPackArray​

class max.driver.DLPackArray(*args, **kwargs)

Device​

class max.driver.Device(_device: Device)

A hardware device abstraction for computation.

Provides a unified interface for CPU and accelerator (e.g. GPU) devices.

accelerator()​

classmethod accelerator(id: int = -1) → Device

Creates an accelerator device with the specified ID.

Provides access to GPU or other hardware accelerators in the system.

from max import driver

# Create default accelerator (usually first available GPU)
device = driver.Accelerator()

# Or specify GPU id
device = driver.Accelerator(id=0)  # First GPU
device = driver.Accelerator(id=1)  # Second GPU

# Get device id
device_id = device.id
from max import driver

# Create default accelerator (usually first available GPU)
device = driver.Accelerator()

# Or specify GPU id
device = driver.Accelerator(id=0)  # First GPU
device = driver.Accelerator(id=1)  # Second GPU

# Get device id
device_id = device.id

api​

property api*: str*

Provides the programming interface used by the device.

  • Returns:

    The programming interface identifier used by the device. One of:

    • cpu for host devices.
    • cuda for NVIDIA GPUs.
    • hip for AMD GPUs.
from max import driver

device = driver.CPU()
device.api
from max import driver

device = driver.CPU()
device.api

cpu()​

classmethod cpu(id: int = -1) → Device

Creates a CPU device for the specified NUMA node.

from max import driver

# Create default CPU device
device = driver.CPU()

# Or specify NUMA node id if using NUMA architecture
device = driver.CPU(id=0)  # First NUMA node
device = driver.CPU(id=1)  # Second NUMA node

# Get device id
device_id = device.id
from max import driver

# Create default CPU device
device = driver.CPU()

# Or specify NUMA node id if using NUMA architecture
device = driver.CPU(id=0)  # First NUMA node
device = driver.CPU(id=1)  # Second NUMA node

# Get device id
device_id = device.id

id​

property id*: int*

Provides the unique identifier for this device.

For CPU devices, this represents the NUMA node ID. For GPU accelerators, this is the device ID relative to the host. Combined with label, this forms a unique device identifier (e.g., gpu:0 or gpu:1).

from max import driver

device = driver.CPU()
device.id
from max import driver

device = driver.CPU()
device.id

is_compatible​

property is_compatible*: bool*

Returns whether this device is compatible with MAX.

is_host​

property is_host

Checks whether this device is the CPU (host) device.

from max import driver

device = driver.CPU()
device.is_host
from max import driver

device = driver.CPU()
device.is_host

label​

property label*: str*

Provides the device type identifier.

from max import driver

device = driver.CPU()
device.label
from max import driver

device = driver.CPU()
device.label

stats​

property stats*: Mapping[str, Any]*

Provides real-time utilization data for the device.

from max import driver

device = driver.CPU()
device.stats
from max import driver

device = driver.CPU()
device.stats

DeviceSpec​

class max.driver.DeviceSpec(id: int, device_type: Literal['cpu', 'gpu'] = 'cpu')

Specification for a device, containing its ID and type.

This class provides a way to specify device parameters like ID and type (CPU/GPU) for creating Device instances.

accelerator()​

static accelerator(id: int = -1)

Creates an accelerator (GPU) device specification.

cpu()​

static cpu(id: int = -1)

Creates a CPU device specification.

device_type​

device_type*: Literal['cpu', 'gpu']* = 'cpu'

Type of specified device.

id​

id*: int*

Provided id for this device.

MemMapTensor​

class max.driver.MemMapTensor(filename: PathLike, dtype: DType, shape: Sequence[int], mode='r+', offset=0)

Create a memory-mapped tensor from a binary file on disk.

The constructor argument semantics follow that of np.memmap.

read_only​

property read_only*: bool*

Tensor​

class max.driver.Tensor(shape: ~typing.Sequence[int], dtype: ~max.dtype.dtype.DType, device: ~max.driver.driver.Device = Device(_device=<max._driver.Device object>))

Device-resident tensor representation. Allocates memory onto a given device with the provided shape and dtype. Tensors can be sliced to provide strided views of the underlying memory, but any tensors input into model execution must be contiguous. Does not currently support setting items across multiple indices, but does support numpy-style slicing.

  • Parameters:

    • dtype – DType of tensor
    • shape – Tuple of positive, non-zero integers denoting the tensor shape.
    • device – Device to allocate tensor onto.

contiguous()​

contiguous() → Tensor

Creates a contiguous copy of the parent tensor.

copy()​

copy(device: Device | None = None) → Tensor

Create a deep copy on an optionally given device.

If a device is None (default), a copy is created on the same device.

from max import driver
from max.dtype import DType

cpu_tensor = driver.Tensor([2, 3], dtype=DType.bfloat16, device=driver.CPU())

cpu_copy = cpu_tensor.copy()
from max import driver
from max.dtype import DType

cpu_tensor = driver.Tensor([2, 3], dtype=DType.bfloat16, device=driver.CPU())

cpu_copy = cpu_tensor.copy()

device​

property device*: Device*

Device on which tensor is resident.

dtype​

property dtype*: DType*

DType of constituent elements in tensor.

element_size​

property element_size*: int*

Return the size of the element type in bytes.

from_dlpack()​

classmethod from_dlpack(arr: Any, *, copy: bool | None = None) → Tensor

Create a tensor from an object implementing the dlpack protocol.

This usually does not result in a copy, and the producer of the object retains ownership of the underlying memory.

from_numpy()​

classmethod from_numpy(arr: ndarray) → Tensor

Creates a tensor from a provided numpy array on the host device.

The underlying data is not copied unless the array is noncontiguous. If it is, a contiguous copy will be returned.

is_contiguous​

property is_contiguous*: bool*

Whether or not tensor is contiguously allocated in memory. Returns false if the tensor is a non-contiguous slice.

Currently, we consider certain situations that are contiguous as non-contiguous for the purposes of our engine, such as when a tensor has negative steps.

is_host​

property is_host*: bool*

Whether or not tensor is host-resident. Returns false for GPU tensors, true for CPU tensors.

from max import driver
from max.dtype import DType

cpu_tensor = driver.Tensor([2, 3], dtype=DType.bfloat16, device=driver.CPU())

print(cpu_tensor.is_host)
from max import driver
from max.dtype import DType

cpu_tensor = driver.Tensor([2, 3], dtype=DType.bfloat16, device=driver.CPU())

print(cpu_tensor.is_host)

item()​

item() → Any

Returns the scalar value at a given location. Currently implemented only for zero-rank tensors. The return type is converted to a Python built-in type.

num_elements​

property num_elements*: int*

Returns the number of elements in this tensor.

Rank-0 tensors have 1 element by convention.

rank​

property rank*: int*

Tensor rank.

scalar()​

classmethod scalar(value: ~typing.Any, dtype: ~max.dtype.dtype.DType, device: ~max.driver.driver.Device = Device(_device=<max._driver.Device object>)) → Tensor

Create a scalar value of a given dtype and value.

shape​

property shape*: Sequence[int]*

Shape of tensor.

to()​

to(device: Device) → Tensor

Return a tensor that’s guaranteed to be on the given device.

The tensor is only copied if the input device is different from the device upon which the tensor is already resident.

to_numpy()​

to_numpy() → ndarray

Converts the tensor to a numpy array.

If the tensor is not on the host, an exception is raised.

view()​

view(dtype: DType, shape: Sequence[int] | None = None) → Tensor

Return a new tensor with the given type and shape that shares the underlying memory.

If the shape is not given, it will be deduced if possible, or a ValueError is raised.

zeros()​

classmethod zeros(shape: ~typing.Sequence[int], dtype: ~max.dtype.dtype.DType, device: ~max.driver.driver.Device = Device(_device=<max._driver.Device object>)) → Tensor

Allocates an tensor with all elements initialized to zero.