MCPcopy Create free account
hub / github.com/apache/tvm / warp_size

Method warp_size

python/tvm/runtime/device.py:70–84  ·  view source on GitHub ↗

Number of threads that execute concurrently. Returns device value for CUDA, ROCm, and Vulkan. Returns 1 for Metal and OpenCL devices, regardless of the physical device. Returns remote device value for RPC devices. Returns None for all other devices. Retur

(self)

Source from the content-addressed store, hash-verified

68
69 @property
70 def warp_size(self):
71 """Number of threads that execute concurrently.
72
73 Returns device value for CUDA, ROCm, and Vulkan. Returns
74 1 for Metal and OpenCL devices, regardless of the physical
75 device. Returns remote device value for RPC devices. Returns
76 None for all other devices.
77
78 Returns
79 -------
80 warp_size : int or None
81 Number of threads that execute concurrently
82
83 """
84 return self._GetDeviceAttr(self.dlpack_device_type(), self.index, 2)
85
86 @property
87 def max_shared_memory_per_block(self):

Callers

nothing calls this directly

Calls 1

_GetDeviceAttrMethod · 0.95

Tested by

no test coverage detected