warp.MemoryKind#
- class warp.MemoryKind(*values)[source]#
Observed memory kind backing a Warp array.
This describes the memory class reported by Warp and CUDA pointer attributes. It does not describe current physical residency, migration state, synchronization state, or whether a specific device can access the array. Use
can_access()for access checks.Methods
conjugateReturns self, the complex conjugate of any int.
bit_length()Number of bits necessary to represent self in binary.
bit_count()Number of ones in the binary representation of the absolute value of self.
to_bytes([length, byteorder, signed])Return an array of bytes representing an integer.
from_bytes(bytes[, byteorder, signed])as_integer_ratio()Return a pair of integers, whose ratio is equal to the original int.
is_integer()Returns True.
__init__(*args, **kwds)Attributes
realthe real part of a complex number
imagthe imaginary part of a complex number
numeratorthe numerator of a rational number in lowest terms
denominatorthe denominator of a rational number in lowest terms
Memory kind that Warp cannot classify.
Default host memory.
Pinned or CUDA-registered host memory.
CUDA device memory that is not classified as managed or mempool memory.
CUDA memory-pool allocation.
CUDA managed-memory allocation.
- UNKNOWN = 0#
Memory kind that Warp cannot classify.
- HOST = 1#
Default host memory.
- PINNED = 2#
Pinned or CUDA-registered host memory.
- CUDA_DEVICE = 3#
CUDA device memory that is not classified as managed or mempool memory.
- CUDA_MEMPOOL = 4#
CUDA memory-pool allocation.
- CUDA_MANAGED = 5#
CUDA managed-memory allocation.