tilelang.carver.arch.driver.cuda_driver¶

Classes¶

Functions¶

get_cuda_device_properties([device_id])

get_device_name([device_id])

get_shared_memory_per_block([device_id, format])

get_device_attribute(attr[, device_id])

get_max_dynamic_shared_size_bytes([device_id, format])

Get the maximum dynamic shared memory size in bytes, kilobytes, or megabytes.

get_persisting_l2_cache_max_size([device_id])

get_num_sms([device_id])

Get the number of streaming multiprocessors (SMs) on the CUDA device.

get_registers_per_block([device_id])

Module Contents¶

class tilelang.carver.arch.driver.cuda_driver.cudaDeviceProp¶

Bases: ctypes.Structure

tilelang.carver.arch.driver.cuda_driver.get_cuda_device_properties(device_id=0)¶
Parameters:

device_id (int)

Return type:

Optional[cudaDeviceProp]

tilelang.carver.arch.driver.cuda_driver.get_device_name(device_id=0)¶
Parameters:

device_id (int)

Return type:

Optional[str]

tilelang.carver.arch.driver.cuda_driver.get_shared_memory_per_block(device_id=0, format='bytes')¶
Parameters:
  • device_id (int)

  • format (str)

Return type:

Optional[int]

tilelang.carver.arch.driver.cuda_driver.get_device_attribute(attr, device_id=0)¶
Parameters:
  • attr (int)

  • device_id (int)

Return type:

int

tilelang.carver.arch.driver.cuda_driver.get_max_dynamic_shared_size_bytes(device_id=0, format='bytes')¶

Get the maximum dynamic shared memory size in bytes, kilobytes, or megabytes.

Parameters:
  • device_id (int)

  • format (str)

Return type:

Optional[int]

tilelang.carver.arch.driver.cuda_driver.get_persisting_l2_cache_max_size(device_id=0)¶
Parameters:

device_id (int)

Return type:

int

tilelang.carver.arch.driver.cuda_driver.get_num_sms(device_id=0)¶

Get the number of streaming multiprocessors (SMs) on the CUDA device.

Parameters:

device_id (int, optional) – The CUDA device ID. Defaults to 0.

Returns:

The number of SMs on the device.

Return type:

int

Raises:

RuntimeError – If unable to get the device properties.

tilelang.carver.arch.driver.cuda_driver.get_registers_per_block(device_id=0)¶
Parameters:

device_id (int)

Return type:

int