CuPy – NumPy-like API accelerated with CUDA

This is the CuPy documentation.

Overview

CuPy is an implementation of NumPy-compatible multi-dimensional array on CUDA. CuPy consists of cupy.ndarray, the core multi-dimensional array class, and many functions on it. It supports a subset of numpy.ndarray interface.

The following is a brief overview of supported subset of NumPy interface:

• Basic indexing (indexing by ints, slices, newaxes, and Ellipsis)
• Most of Advanced indexing (except for some indexing patterns with boolean masks)
• Data types (dtypes): bool_, int8, int16, int32, int64, uint8, uint16, uint32, uint64, float16, float32, float64, complex64, complex128
• Most of the array creation routines (empty, ones_like, diag, etc.)
• Most of the array manipulation routines (reshape, rollaxis, concatenate, etc.)
• All operators with broadcasting
• All universal functions for elementwise operations (except those for complex numbers).
• Linear algebra functions, including product (dot, matmul, etc.) and decomposition (cholesky, svd, etc.), accelerated by cuBLAS.
• Reduction along axes (sum, max, argmax, etc.)

CuPy also includes the following features for performance:

• User-defined elementwise CUDA kernels
• User-defined reduction CUDA kernels
• Fusing CUDA kernels to optimize user-defined calculation
• Customizable memory allocator and memory pool
• cuDNN utilities

CuPy uses on-the-fly kernel synthesis: when a kernel call is required, it compiles a kernel code optimized for the shapes and dtypes of given arguments, sends it to the GPU device, and executes the kernel. The compiled code is cached to $(HOME)/.cupy/kernel_cache directory (this cache path can be overwritten by setting the CUPY_CACHE_DIR environment variable). It may make things slower at the first kernel call, though this slow down will be resolved at the second execution. CuPy also caches the kernel code sent to GPU device within the process, which reduces the kernel transfer time on further calls.

Tutorial

Basics of CuPy

In this section, you will learn about the following things:

• Basics of cupy.ndarray
• The concept of current device
• host-device and device-device array transfer

Basics of cupy.ndarray

CuPy is a GPU array backend that implements a subset of NumPy interface. In the following code, cp is an abbreviation of cupy, as np is numpy as is customarily done:

>>> import numpy as np
>>> import cupy as cp

The cupy.ndarray class is in its core, which is a compatible GPU alternative of numpy.ndarray.

>>> x_gpu = cp.array([1, 2, 3])

x_gpu in the above example is an instance of cupy.ndarray. You can see its creation of identical to NumPy’s one, except that numpy is replaced with cupy. The main difference of cupy.ndarray from numpy.ndarray is that the content is allocated on the device memory. Its data is allocated on the current device, which will be explained later.

Most of the array manipulations are also done in the way similar to NumPy. Take the Euclidean norm (a.k.a L2 norm) for example. NumPy has numpy.linalg.norm() to calculate it on CPU.

>>> x_cpu = np.array([1, 2, 3])
>>> l2_cpu = np.linalg.norm(x_cpu)

We can calculate it on GPU with CuPy in a similar way:

>>> x_gpu = cp.array([1, 2, 3])
>>> l2_gpu = cp.linalg.norm(x_gpu)

CuPy implements many functions on cupy.ndarray objects. See the reference for the supported subset of NumPy API. Understanding NumPy might help utilizing most features of CuPy. So, we recommend you to read the NumPy documentation.

Current Device

CuPy has a concept of the current device, which is the default device on which the allocation, manipulation, calculation etc. of arrays are taken place. Suppose the ID of current device is 0. The following code allocates array contents on GPU 0.

>>> x_on_gpu0 = cp.array([1, 2, 3, 4, 5])

The current device can be changed by cupy.cuda.Device.use() as follows:

>>> x_on_gpu0 = cp.array([1, 2, 3, 4, 5])
>>> cp.cuda.Device(1).use()
>>> x_on_gpu1 = cp.array([1, 2, 3, 4, 5])

If you switch the current GPU temporarily, with statement comes in handy.

>>> with cp.cuda.Device(1):
...    x_on_gpu1 = cp.array([1, 2, 3, 4, 5])
>>> x_on_gpu0 = cp.array([1, 2, 3, 4, 5])

Most operations of CuPy is done on the current device. Be careful that if processing of an array on a non-current device will cause an error:

>>> with cp.cuda.Device(0):
...    x_on_gpu0 = cp.array([1, 2, 3, 4, 5])
>>> with cp.cuda.Device(1):
...    x_on_gpu0 * 2  # raises error
Traceback (most recent call last):
...
ValueError: Array device must be same as the current device: array device = 0 while current = 1

cupy.ndarray.device attribute indicates the device on which the array is allocated.

>>> with cp.cuda.Device(1):
...    x = cp.array([1, 2, 3, 4, 5])
>>> x.device
<CUDA Device 1>

Note

If the environment has only one device, such explicit device switching is not needed.

Data Transfer

Move arrays to a device

cupy.asarray() can be used to move a numpy.ndarray, a list, or any object that can be passed to numpy.array() to the current device:

>>> x_cpu = np.array([1, 2, 3])
>>> x_gpu = cp.asarray(x_cpu)  # move the data to the current device.

cupy.asarray() can accept cupy.ndarray, which means we can transfer the array between devices with this function.

>>> with cp.cuda.Device(0):
...     x_gpu_0 = cp.ndarray([1, 2, 3])  # create an array in GPU 0
>>> with cp.cuda.Device(1):
...     x_gpu_1 = cp.asarray(x_gpu_0)  # move the array to GPU 1

Note

cupy.asarray() does not copy the input array if possible. So, if you put an array of the current device, it returns the input object itself.

If we do copy the array in this situation, you can use cupy.array() with copy=True. Actually cupy.asarray() is equivalent to cupy.array(arr, dtype, copy=False).

Move array from a device to the host

Moving a device array to the host can be done by cupy.asnumpy() as follows:

>>> x_gpu = cp.array([1, 2, 3])  # create an array in the current device
>>> x_cpu = cp.asnumpy(x_gpu)  # move the array to the host.

We can also use cupy.ndarray.get():

>>> x_cpu = x_gpu.get()

Note

If you work with Chainer, you can also use to_cpu() and to_gpu() to move arrays back and forth between a device and a host, or between different devices. Note that to_gpu() has device option to specify the device which arrays are transferred.

How to write CPU/GPU agnostic code

The compatibility of CuPy with NumPy enables us to write CPU/GPU generic code. It can be made easy by the cupy.get_array_module() function. This function returns the numpy or cupy module based on arguments. A CPU/GPU generic function is defined using it like follows:

>>> # Stable implementation of log(1 + exp(x))
>>> def softplus(x):
...     xp = cp.get_array_module(x)
...     return xp.maximum(0, x) + xp.log1p(xp.exp(-abs(x)))

Sometimes, an explicit conversion to a host or device array may be required. cupy.asarray() and cupy.asnumpy() can be used in agnostic implementations to get host or device arrays from either CuPy or NumPy arrays.

>>> y_cpu = np.array([4, 5, 6])
>>> x_cpu + y_cpu
array([5, 7, 9])
>>> x_gpu + y_cpu
Traceback (most recent call last):
...
TypeError: Unsupported type <class 'numpy.ndarray'>
>>> cp.asnumpy(x_gpu) + y_cpu
array([5, 7, 9])
>>> cp.asnumpy(x_gpu) + cp.asnumpy(y_cpu)
array([5, 7, 9])
>>> x_gpu + cp.asarray(y_cpu)
array([5, 7, 9])
>>> cp.asarray(x_gpu) + cp.asarray(y_cpu)
array([5, 7, 9])

User-Defined Kernels

CuPy provides easy ways to define three types of CUDA kernels: elementwise kernels, reduction kernels and raw kernels. In this documentation, we describe how to define and call each kernels.

Basics of elementwise kernels

An elementwise kernel can be defined by the ElementwiseKernel class. The instance of this class defines a CUDA kernel which can be invoked by the __call__ method of this instance.

A definition of an elementwise kernel consists of four parts: an input argument list, an output argument list, a loop body code, and the kernel name. For example, a kernel that computes a squared difference \(f(x, y) = (x - y)^2\) is defined as follows:

>>> squared_diff = cp.ElementwiseKernel(
...    'float32 x, float32 y',
...    'float32 z',
...    'z = (x - y) * (x - y)',
...    'squared_diff')

The argument lists consist of comma-separated argument definitions. Each argument definition consists of a type specifier and an argument name. Names of NumPy data types can be used as type specifiers.

Note

n, i, and names starting with an underscore _ are reserved for the internal use.

The above kernel can be called on either scalars or arrays with broadcasting:

>>> x = cp.arange(10, dtype=np.float32).reshape(2, 5)
>>> y = cp.arange(5, dtype=np.float32)
>>> squared_diff(x, y)
array([[ 0.,  0.,  0.,  0.,  0.],
       [25., 25., 25., 25., 25.]], dtype=float32)
>>> squared_diff(x, 5)
array([[25., 16.,  9.,  4.,  1.],
       [ 0.,  1.,  4.,  9., 16.]], dtype=float32)

Output arguments can be explicitly specified (next to the input arguments):

>>> z = cp.empty((2, 5), dtype=np.float32)
>>> squared_diff(x, y, z)
array([[ 0.,  0.,  0.,  0.,  0.],
       [25., 25., 25., 25., 25.]], dtype=float32)

Type-generic kernels

If a type specifier is one character, then it is treated as a type placeholder. It can be used to define a type-generic kernels. For example, the above squared_diff kernel can be made type-generic as follows:

>>> squared_diff_generic = cp.ElementwiseKernel(
...     'T x, T y',
...     'T z',
...     'z = (x - y) * (x - y)',
...     'squared_diff_generic')

Type placeholders of a same character in the kernel definition indicate the same type. The actual type of these placeholders is determined by the actual argument type. The ElementwiseKernel class first checks the output arguments and then the input arguments to determine the actual type. If no output arguments are given on the kernel invocation, then only the input arguments are used to determine the type.

The type placeholder can be used in the loop body code:

>>> squared_diff_generic = cp.ElementwiseKernel(
...     'T x, T y',
...     'T z',
...     '''
...         T diff = x - y;
...         z = diff * diff;
...     ''',
...     'squared_diff_generic')

More than one type placeholder can be used in a kernel definition. For example, the above kernel can be further made generic over multiple arguments:

>>> squared_diff_super_generic = cp.ElementwiseKernel(
...     'X x, Y y',
...     'Z z',
...     'z = (x - y) * (x - y)',
...     'squared_diff_super_generic')

Note that this kernel requires the output argument explicitly specified, because the type Z cannot be automatically determined from the input arguments.

Raw argument specifiers

The ElementwiseKernel class does the indexing with broadcasting automatically, which is useful to define most elementwise computations. On the other hand, we sometimes want to write a kernel with manual indexing for some arguments. We can tell the ElementwiseKernel class to use manual indexing by adding the raw keyword preceding the type specifier.

We can use the special variable i and method _ind.size() for the manual indexing. i indicates the index within the loop. _ind.size() indicates total number of elements to apply the elementwise operation. Note that it represents the size after broadcast operation.

For example, a kernel that adds two vectors with reversing one of them can be written as follows:

>>> add_reverse = cp.ElementwiseKernel(
...     'T x, raw T y', 'T z',
...     'z = x + y[_ind.size() - i - 1]',
...     'add_reverse')

(Note that this is an artificial example and you can write such operation just by z = x + y[::-1] without defining a new kernel). A raw argument can be used like an array. The indexing operator y[_ind.size() - i - 1] involves an indexing computation on y, so y can be arbitrarily shaped and strode.

Note that raw arguments are not involved in the broadcasting. If you want to mark all arguments as raw, you must specify the size argument on invocation, which defines the value of _ind.size().

Reduction kernels

Reduction kernels can be defined by the ReductionKernel class. We can use it by defining four parts of the kernel code:

1. Identity value: This value is used for the initial value of reduction.
2. Mapping expression: It is used for the pre-processing of each element to be reduced.
3. Reduction expression: It is an operator to reduce the multiple mapped values. The special variables a and b are used for its operands.
4. Post mapping expression: It is used to transform the resulting reduced values. The special variable a is used as its input. Output should be written to the output parameter.

ReductionKernel class automatically inserts other code fragments that are required for an efficient and flexible reduction implementation.

For example, L2 norm along specified axes can be written as follows:

>>> l2norm_kernel = cp.ReductionKernel(
...     'T x',  # input params
...     'T y',  # output params
...     'x * x',  # map
...     'a + b',  # reduce
...     'y = sqrt(a)',  # post-reduction map
...     '0',  # identity value
...     'l2norm'  # kernel name
... )
>>> x = cp.arange(10, dtype=np.float32).reshape(2, 5)
>>> l2norm_kernel(x, axis=1)
array([ 5.477226 , 15.9687195], dtype=float32)

Note

raw specifier is restricted for usages that the axes to be reduced are put at the head of the shape. It means, if you want to use raw specifier for at least one argument, the axis argument must be 0 or a contiguous increasing sequence of integers starting from 0, like (0, 1), (0, 1, 2), etc.

Raw kernels

Raw kernels can be defined by the RawKernel class. By using raw kernels, you can define kernels from raw CUDA source.

RawKernel object allows you to call the kernel with CUDA’s cuLaunchKernel interface. In other words, you have control over grid size, block size, shared memory size and stream.

>>> add_kernel = cp.RawKernel(r'''
... extern "C" __global__
... void my_add(const float* x1, const float* x2, float* y) {
...     int tid = blockDim.x * blockIdx.x + threadIdx.x;
...     y[tid] = x1[tid] + x2[tid];
... }
... ''', 'my_add')
>>> x1 = cp.arange(25, dtype=cp.float32).reshape(5, 5)
>>> x2 = cp.arange(25, dtype=cp.float32).reshape(5, 5)
>>> y = cp.zeros((5, 5), dtype=cp.float32)
>>> add_kernel((5,), (5,), (x1, x2, y))  # grid, block and arguments
>>> y
array([[ 0.,  2.,  4.,  6.,  8.],
       [10., 12., 14., 16., 18.],
       [20., 22., 24., 26., 28.],
       [30., 32., 34., 36., 38.],
       [40., 42., 44., 46., 48.]], dtype=float32)

Raw kernels operating on complex-valued arrays can be created as well:

>>> complex_kernel = cp.RawKernel(r'''
... #include <cupy/complex.cuh>
... extern "C" __global__
... void my_func(const complex<float>* x1, const complex<float>* x2,
...              complex<float>* y, float a) {
...     int tid = blockDim.x * blockIdx.x + threadIdx.x;
...     y[tid] = x1[tid] + a * x2[tid];
... }
... ''', 'my_func')
>>> x1 = cupy.arange(25, dtype=cupy.complex64).reshape(5, 5)
>>> x2 = 1j*cupy.arange(25, dtype=cupy.complex64).reshape(5, 5)
>>> y = cupy.zeros((5, 5), dtype=cupy.complex64)
>>> complex_kernel((5,), (5,), (x1, x2, y, cupy.float32(2.0)))  # grid, block and arguments
>>> y
array([[ 0. +0.j,  1. +2.j,  2. +4.j,  3. +6.j,  4. +8.j],
       [ 5.+10.j,  6.+12.j,  7.+14.j,  8.+16.j,  9.+18.j],
       [10.+20.j, 11.+22.j, 12.+24.j, 13.+26.j, 14.+28.j],
       [15.+30.j, 16.+32.j, 17.+34.j, 18.+36.j, 19.+38.j],
       [20.+40.j, 21.+42.j, 22.+44.j, 23.+46.j, 24.+48.j]],
      dtype=complex64)

Note that while we encourage the usage of complex<T> types for complex numbers (available by including <cupy/complex.cuh> as shown above), for CUDA codes already written using functions from cuComplex.h there is no need to make the conversion yourself: just set the option translate_cucomplex=True when creating a RawKernel instance.

The CUDA kernel attributes can be retrieved by either accessing the attributes dictionary, or by accessing the RawKernel object’s attributes directly; the latter can also be used to set certain attributes:

>>> add_kernel = cp.RawKernel(r'''
... extern "C" __global__
... void my_add(const float* x1, const float* x2, float* y) {
...     int tid = blockDim.x * blockIdx.x + threadIdx.x;
...     y[tid] = x1[tid] + x2[tid];
... }
... ''', 'my_add')
>>> add_kernel.attributes  
{'max_threads_per_block': 1024, 'shared_size_bytes': 0, 'const_size_bytes': 0, 'local_size_bytes': 0, 'num_regs': 10, 'ptx_version': 70, 'binary_version': 70, 'cache_mode_ca': 0, 'max_dynamic_shared_size_bytes': 49152, 'preferred_shared_memory_carveout': -1}
>>> add_kernel.max_dynamic_shared_size_bytes  
49152
>>> add_kernel.max_dynamic_shared_size_bytes = 50000  # set a new value for the attribute  
>>> add_kernel.max_dynamic_shared_size_bytes  
50000

Dynamical parallelism is supported by RawKernel. You just need to provide the linking flag (such as -dc) to RawKernel’s options arugment. The static CUDA device runtime library (cudadevrt) is automatically discovered by CuPy. For further detail, see CUDA Toolkit’s documentation.

Accessing texture memory in RawKernel is supported via CUDA Runtime’s Texture Object API, see TextureObject’s documentation as well as CUDA C Programming Guide. For using the Texture Reference API, which is marked as deprecated as of CUDA Toolkit 10.1, see the introduction to RawModule below.

Note

The kernel does not have return values. You need to pass both input arrays and output arrays as arguments.

Note

No validation will be performed by CuPy for arguments passed to the kernel, including types and number of arguments. Especially note that when passing ndarray, its dtype should match with the type of the argument declared in the method signature of the CUDA source code (unless you are casting arrays intentionally). For example, cupy.float32 and cupy.uint64 arrays must be passed to the argument typed as float* and unsigned long long*. For Python primitive types, int, float and bool map to long long, double and bool, respectively.

Note

When using printf() in your CUDA kernel, you may need to synchronize the stream to see the output. You can use cupy.cuda.Stream.null.synchronize() if you are using the default stream.

Raw modules

For dealing a large raw CUDA source or loading an existing CUDA binary, the RawModule class can be more handy. It can be initialized either by a CUDA source code, or by a path to the CUDA binary. The needed kernels can then be retrieved by calling the get_function() method, which returns a RawKernel instance that can be invoked as discussed above.

>>> loaded_from_source = r'''
... extern "C"{
...
... __global__ void test_sum(const float* x1, const float* x2, float* y, \
...                          unsigned int N)
... {
...     unsigned int tid = blockDim.x * blockIdx.x + threadIdx.x;
...     if (tid < N)
...     {
...         y[tid] = x1[tid] + x2[tid];
...     }
... }
...
... __global__ void test_multiply(const float* x1, const float* x2, float* y, \
...                               unsigned int N)
... {
...     unsigned int tid = blockDim.x * blockIdx.x + threadIdx.x;
...     if (tid < N)
...     {
...         y[tid] = x1[tid] * x2[tid];
...     }
... }
...
... }'''
>>> module = cp.RawModule(code=loaded_from_source)
>>> ker_sum = module.get_function('test_sum')
>>> ker_times = module.get_function('test_multiply')
>>> N = 10
>>> x1 = cp.arange(N**2, dtype=cp.float32).reshape(N, N)
>>> x2 = cp.ones((N, N), dtype=cp.float32)
>>> y = cp.zeros((N, N), dtype=cp.float32)
>>> ker_sum((N,), (N,), (x1, x2, y, N**2))   # y = x1 + x2
>>> assert cp.allclose(y, x1 + x2)
>>> ker_times((N,), (N,), (x1, x2, y, N**2)) # y = x1 * x2
>>> assert cp.allclose(y, x1 * x2)

The instruction above for using complex numbers in RawKernel also applies to RawModule.

CuPy also supports the Texture Reference API. A handle to the texture reference in a module can be retrieved by name via get_texref(). Then, you need to pass it to TextureReference, along with a resource descriptor and texture descriptor, for binding the reference to the array. (The interface of TextureReference is meant to mimic that of TextureObject to help users make transition to the latter, since as of CUDA Toolkit 10.1 the former is marked as deprecated.)

Kernel fusion

cupy.fuse() is a decorator that fuses functions. This decorator can be used to define an elementwise or reduction kernel more easily than ElementwiseKernel or ReductionKernel.

By using this decorator, we can define the squared_diff kernel as follows:

>>> @cp.fuse()
... def squared_diff(x, y):
...     return (x - y) * (x - y)

The above kernel can be called on either scalars, NumPy arrays or CuPy arrays likes the original function.

>>> x_cp = cp.arange(10)
>>> y_cp = cp.arange(10)[::-1]
>>> squared_diff(x_cp, y_cp)
array([81, 49, 25,  9,  1,  1,  9, 25, 49, 81])
>>> x_np = np.arange(10)
>>> y_np = np.arange(10)[::-1]
>>> squared_diff(x_np, y_np)
array([81, 49, 25,  9,  1,  1,  9, 25, 49, 81])

At the first function call, the fused function analyzes the original function based on the abstracted information of arguments (e.g. their dtypes and ndims) and creates and caches an actual CUDA kernel. From the second function call with the same input types, the fused function calls the previously cached kernel, so it is highly recommended to reuse the same decorated functions instead of decorating local functions that are defined multiple times.

cupy.fuse() also supports simple reduction kernel.

>>> @cp.fuse()
... def sum_of_products(x, y):
...     return cp.sum(x * y, axis = -1)

You can specify the kernel name by using the kernel_name keyword argument as follows:

>>> @cp.fuse(kernel_name='squared_diff')
... def squared_diff(x, y):
...     return (x - y) * (x - y)

Note

Currently, cupy.fuse() can fuse only simple elementwise and reduction operations. Most other routines (e.g. cupy.matmul(), cupy.reshape()) are not supported.

Reference Manual

This is the official reference of CuPy, a multi-dimensional array on CUDA with a subset of NumPy interface.

• Index
• Module Index

---

Multi-Dimensional Array (ndarray)

cupy.ndarray is the CuPy counterpart of NumPy numpy.ndarray. It provides an intuitive interface for a fixed-size multidimensional array which resides in a CUDA device.

For the basic concept of ndarrays, please refer to the NumPy documentation.

cupy.ndarray

Multi-dimensional array on a CUDA device.

Code compatibility features

cupy.ndarray is designed to be interchangeable with numpy.ndarray in terms of code compatibility as much as possible. But occasionally, you will need to know whether the arrays you’re handling are cupy.ndarray or numpy.ndarray. One example is when invoking module-level functions such as cupy.sum() or numpy.sum(). In such situations, cupy.get_array_module() can be used.

cupy.get_array_module	Returns the array module for arguments.
cupyx.scipy.get_array_module	Returns the array module for arguments.

Conversion to/from NumPy arrays

cupy.ndarray and numpy.ndarray are not implicitly convertible to each other. That means, NumPy functions cannot take cupy.ndarrays as inputs, and vice versa.

• To convert numpy.ndarray to cupy.ndarray, use cupy.array() or cupy.asarray().
• To convert cupy.ndarray to numpy.ndarray, use cupy.asnumpy() or cupy.ndarray.get().

Note that converting between cupy.ndarray and numpy.ndarray incurs data transfer between the host (CPU) device and the GPU device, which is costly in terms of performance.

cupy.array	Creates an array on the current device.
cupy.asarray	Converts an object to array.
cupy.asnumpy	Returns an array on the host memory from an arbitrary source array.

Universal Functions (ufunc)

CuPy provides universal functions (a.k.a. ufuncs) to support various elementwise operations. CuPy’s ufunc supports following features of NumPy’s one:

• Broadcasting
• Output type determination
• Casting rules

CuPy’s ufunc currently does not provide methods such as reduce, accumulate, reduceat, outer, and at.

Ufunc class

cupy.ufunc

Universal function.

Available ufuncs

Math operations

cupy.add	Adds two arrays elementwise.
cupy.subtract	Subtracts arguments elementwise.
cupy.multiply	Multiplies two arrays elementwise.
cupy.divide	Elementwise true division (i.e.
cupy.logaddexp	Computes log(exp(x1) + exp(x2)) elementwise.
cupy.logaddexp2	Computes log2(exp2(x1) + exp2(x2)) elementwise.
cupy.true_divide	Elementwise true division (i.e.
cupy.floor_divide	Elementwise floor division (i.e.
cupy.negative	Takes numerical negative elementwise.
cupy.power	Computes x1 ** x2 elementwise.
cupy.remainder	Computes the remainder of Python division elementwise.
cupy.mod	Computes the remainder of Python division elementwise.
cupy.fmod	Computes the remainder of C division elementwise.
cupy.absolute	Elementwise absolute value function.
cupy.rint	Rounds each element of an array to the nearest integer.
cupy.sign	Elementwise sign function.
cupy.exp	Elementwise exponential function.
cupy.exp2	Elementwise exponentiation with base 2.
cupy.log	Elementwise natural logarithm function.
cupy.log2	Elementwise binary logarithm function.
cupy.log10	Elementwise common logarithm function.
cupy.expm1	Computes exp(x) - 1 elementwise.
cupy.log1p	Computes log(1 + x) elementwise.
cupy.sqrt	Elementwise square root function.
cupy.square	Elementwise square function.
cupy.reciprocal	Computes 1 / x elementwise.

Trigonometric functions

cupy.sin	Elementwise sine function.
cupy.cos	Elementwise cosine function.
cupy.tan	Elementwise tangent function.
cupy.arcsin	Elementwise inverse-sine function (a.k.a.
cupy.arccos	Elementwise inverse-cosine function (a.k.a.
cupy.arctan	Elementwise inverse-tangent function (a.k.a.
cupy.arctan2	Elementwise inverse-tangent of the ratio of two arrays.
cupy.hypot	Computes the hypoteneous of orthogonal vectors of given length.
cupy.sinh	Elementwise hyperbolic sine function.
cupy.cosh	Elementwise hyperbolic cosine function.
cupy.tanh	Elementwise hyperbolic tangent function.
cupy.arcsinh	Elementwise inverse of hyperbolic sine function.
cupy.arccosh	Elementwise inverse of hyperbolic cosine function.
cupy.arctanh	Elementwise inverse of hyperbolic tangent function.
cupy.deg2rad	Converts angles from degrees to radians elementwise.
cupy.rad2deg	Converts angles from radians to degrees elementwise.

Bit-twiddling functions

cupy.bitwise_and	Computes the bitwise AND of two arrays elementwise.
cupy.bitwise_or	Computes the bitwise OR of two arrays elementwise.
cupy.bitwise_xor	Computes the bitwise XOR of two arrays elementwise.
cupy.invert	Computes the bitwise NOT of an array elementwise.
cupy.left_shift	Shifts the bits of each integer element to the left.
cupy.right_shift	Shifts the bits of each integer element to the right.

Comparison functions

cupy.greater	Tests elementwise if x1 > x2.
cupy.greater_equal	Tests elementwise if x1 >= x2.
cupy.less	Tests elementwise if x1 < x2.
cupy.less_equal	Tests elementwise if x1 <= x2.
cupy.not_equal	Tests elementwise if x1 != x2.
cupy.equal	Tests elementwise if x1 == x2.
cupy.logical_and	Computes the logical AND of two arrays.
cupy.logical_or	Computes the logical OR of two arrays.
cupy.logical_xor	Computes the logical XOR of two arrays.
cupy.logical_not	Computes the logical NOT of an array.
cupy.maximum	Takes the maximum of two arrays elementwise.
cupy.minimum	Takes the minimum of two arrays elementwise.
cupy.fmax	Takes the maximum of two arrays elementwise.
cupy.fmin	Takes the minimum of two arrays elementwise.

Floating functions

cupy.isfinite	Tests finiteness elementwise.
cupy.isinf	Tests if each element is the positive or negative infinity.
cupy.isnan	Tests if each element is a NaN.
cupy.signbit	Tests elementwise if the sign bit is set (i.e.
cupy.copysign	Returns the first argument with the sign bit of the second elementwise.
cupy.nextafter	Computes the nearest neighbor float values towards the second argument.
cupy.modf	Extracts the fractional and integral parts of an array elementwise.
cupy.ldexp	Computes x1 * 2 ** x2 elementwise.
cupy.frexp	Decomposes each element to mantissa and two’s exponent.
cupy.fmod	Computes the remainder of C division elementwise.
cupy.floor	Rounds each element of an array to its floor integer.
cupy.ceil	Rounds each element of an array to its ceiling integer.
cupy.trunc	Rounds each element of an array towards zero.

ufunc.at

Currently, CuPy does not support at for ufuncs in general. However, cupyx.scatter_add() can substitute add.at as both behave identically.

Routines

The following pages describe NumPy-compatible routines. These functions cover a subset of NumPy routines.

Array Creation Routines

Basic creation routines

cupy.empty	Returns an array without initializing the elements.
cupy.empty_like	Returns a new array with same shape and dtype of a given array.
cupy.eye	Returns a 2-D array with ones on the diagonals and zeros elsewhere.
cupy.identity	Returns a 2-D identity array.
cupy.ones	Returns a new array of given shape and dtype, filled with ones.
cupy.ones_like	Returns an array of ones with same shape and dtype as a given array.
cupy.zeros	Returns a new array of given shape and dtype, filled with zeros.
cupy.zeros_like	Returns an array of zeros with same shape and dtype as a given array.
cupy.full	Returns a new array of given shape and dtype, filled with a given value.
cupy.full_like	Returns a full array with same shape and dtype as a given array.

Creation from other data

cupy.array	Creates an array on the current device.
cupy.asarray	Converts an object to array.
cupy.asanyarray	Converts an object to array.
cupy.ascontiguousarray	Returns a C-contiguous array.
cupy.copy	Creates a copy of a given array on the current device.
cupy.fromfile	Reads an array from a file.

Numerical ranges

cupy.arange	Returns an array with evenly spaced values within a given interval.
cupy.linspace	Returns an array with evenly-spaced values within a given interval.
cupy.logspace	Returns an array with evenly-spaced values on a log-scale.
cupy.meshgrid	Return coordinate matrices from coordinate vectors.
cupy.mgrid	Construct a multi-dimensional “meshgrid”.
cupy.ogrid	Construct a multi-dimensional “meshgrid”.

Matrix creation

cupy.diag	Returns a diagonal or a diagonal array.
cupy.diagflat	Creates a diagonal array from the flattened input.
cupy.tri	Creates an array with ones at and below the given diagonal.
cupy.tril	Returns a lower triangle of an array.
cupy.triu	Returns an upper triangle of an array.

Array Manipulation Routines

Basic operations

cupy.copyto

Copies values from one array to another with broadcasting.

Changing array shape

cupy.reshape	Returns an array with new shape and same elements.
cupy.ravel	Returns a flattened array.

Transpose-like operations

cupy.moveaxis	Moves axes of an array to new positions.
cupy.rollaxis	Moves the specified axis backwards to the given place.
cupy.swapaxes	Swaps the two axes.
cupy.transpose	Permutes the dimensions of an array.

See also

cupy.ndarray.T

Changing number of dimensions

cupy.atleast_1d	Converts arrays to arrays with dimensions >= 1.
cupy.atleast_2d	Converts arrays to arrays with dimensions >= 2.
cupy.atleast_3d	Converts arrays to arrays with dimensions >= 3.
cupy.broadcast	Object that performs broadcasting.
cupy.broadcast_to	Broadcast an array to a given shape.
cupy.broadcast_arrays	Broadcasts given arrays.
cupy.expand_dims	Expands given arrays.
cupy.squeeze	Removes size-one axes from the shape of an array.

Changing kind of array

cupy.asarray	Converts an object to array.
cupy.asanyarray	Converts an object to array.
cupy.asfortranarray	Return an array laid out in Fortran order in memory.
cupy.ascontiguousarray	Returns a C-contiguous array.

Joining arrays

cupy.concatenate	Joins arrays along an axis.
cupy.stack	Stacks arrays along a new axis.
cupy.column_stack	Stacks 1-D and 2-D arrays as columns into a 2-D array.
cupy.dstack	Stacks arrays along the third axis.
cupy.hstack	Stacks arrays horizontally.
cupy.vstack	Stacks arrays vertically.

Splitting arrays

cupy.split	Splits an array into multiple sub arrays along a given axis.
cupy.array_split	Splits an array into multiple sub arrays along a given axis.
cupy.dsplit	Splits an array into multiple sub arrays along the third axis.
cupy.hsplit	Splits an array into multiple sub arrays horizontally.
cupy.vsplit	Splits an array into multiple sub arrays along the first axis.

Tiling arrays

cupy.tile	Construct an array by repeating A the number of times given by reps.
cupy.repeat	Repeat arrays along an axis.

Adding and removing elements

cupy.unique

Find the unique elements of an array.

Rearranging elements

cupy.flip	Reverse the order of elements in an array along the given axis.
cupy.fliplr	Flip array in the left/right direction.
cupy.flipud	Flip array in the up/down direction.
cupy.reshape	Returns an array with new shape and same elements.
cupy.roll	Roll array elements along a given axis.
cupy.rot90	Rotate an array by 90 degrees in the plane specified by axes.

Binary Operations

Elementwise bit operations

cupy.bitwise_and	Computes the bitwise AND of two arrays elementwise.
cupy.bitwise_or	Computes the bitwise OR of two arrays elementwise.
cupy.bitwise_xor	Computes the bitwise XOR of two arrays elementwise.
cupy.invert	Computes the bitwise NOT of an array elementwise.
cupy.left_shift	Shifts the bits of each integer element to the left.
cupy.right_shift	Shifts the bits of each integer element to the right.

Bit packing

cupy.packbits	Packs the elements of a binary-valued array into bits in a uint8 array.
cupy.unpackbits	Unpacks elements of a uint8 array into a binary-valued output array.

Output formatting

cupy.binary_repr

Return the binary representation of the input number as a string.

Data Type Routines

cupy.can_cast	Returns True if cast between data types can occur according to the casting rule.
cupy.result_type	Returns the type that results from applying the NumPy type promotion rules to the arguments.
cupy.common_type	Return a scalar type which is common to the input arrays.

cupy.promote_types (alias of numpy.promote_types())

cupy.min_scalar_type (alias of numpy.min_scalar_type())

cupy.obj2sctype (alias of numpy.obj2sctype())

Creating data types

cupy.dtype (alias of numpy.dtype)

cupy.format_parser (alias of numpy.format_parser)

Data type information

cupy.finfo (alias of numpy.finfo)

cupy.iinfo (alias of numpy.iinfo)

cupy.MachAr (alias of numpy.MachAr)

Data type testing

cupy.issctype (alias of numpy.issctype())

cupy.issubdtype (alias of numpy.issubdtype())

cupy.issubsctype (alias of numpy.issubsctype())

cupy.issubclass_ (alias of numpy.issubclass_())

cupy.find_common_type (alias of numpy.find_common_type())

Miscellaneous

cupy.typename (alias of numpy.typename())

cupy.sctype2char (alias of numpy.sctype2char())

cupy.mintypecode (alias of numpy.mintypecode())

FFT Functions

Standard FFTs

cupy.fft.fft	Compute the one-dimensional FFT.
cupy.fft.ifft	Compute the one-dimensional inverse FFT.
cupy.fft.fft2	Compute the two-dimensional FFT.
cupy.fft.ifft2	Compute the two-dimensional inverse FFT.
cupy.fft.fftn	Compute the N-dimensional FFT.
cupy.fft.ifftn	Compute the N-dimensional inverse FFT.

Real FFTs

cupy.fft.rfft	Compute the one-dimensional FFT for real input.
cupy.fft.irfft	Compute the one-dimensional inverse FFT for real input.
cupy.fft.rfft2	Compute the two-dimensional FFT for real input.
cupy.fft.irfft2	Compute the two-dimensional inverse FFT for real input.
cupy.fft.rfftn	Compute the N-dimensional FFT for real input.
cupy.fft.irfftn	Compute the N-dimensional inverse FFT for real input.

Hermitian FFTs

cupy.fft.hfft	Compute the FFT of a signal that has Hermitian symmetry.
cupy.fft.ihfft	Compute the FFT of a signal that has Hermitian symmetry.

Helper routines

cupy.fft.fftfreq	Return the FFT sample frequencies.
cupy.fft.rfftfreq	Return the FFT sample frequencies for real input.
cupy.fft.fftshift	Shift the zero-frequency component to the center of the spectrum.
cupy.fft.ifftshift	The inverse of fftshift().

Normalization

The default normalization has the direct transforms unscaled and the inverse transforms are scaled by \(1/n\). If the ketyword argument norm is "ortho", both transforms will be scaled by \(1/\sqrt{n}\).

Code compatibility features

FFT functions of NumPy alway return numpy.ndarray which type is numpy.complex128 or numpy.float64. CuPy functions do not follow the behavior, they will return numpy.complex64 or numpy.float32 if the type of the input is numpy.float16, numpy.float32, or numpy.complex64.

Internally, cupy.fft always generates a cuFFT plan (see the cuFFT documentation for detail) corresponding to the desired transform. When possible, an n-dimensional plan will be used, as opposed to applying separate 1D plans for each axis to be transformed. Using n-dimensional planning can provide better performance for multidimensional transforms, but requires more GPU memory than separable 1D planning. The user can disable n-dimensional planning by setting cupy.fft.config.enable_nd_planning = False. This ability to adjust the planning type is a deviation from the NumPy API, which does not use precomputed FFT plans.

Moreover, the automatic plan generation can be suppressed by using an existing plan returned by cupyx.scipy.fftpack.get_fft_plan() as a context manager. This is again a deviation from NumPy.

Indexing Routines

cupy.c_
cupy.r_
cupy.nonzero	Return the indices of the elements that are non-zero.
cupy.where	Return elements, either from x or y, depending on condition.
cupy.indices	Returns an array representing the indices of a grid.
cupy.ix_	Construct an open mesh from multiple sequences.
cupy.unravel_index	Converts array of flat indices into a tuple of coordinate arrays.
cupy.take	Takes elements of an array at specified indices along an axis.
cupy.take_along_axis	Take values from the input array by matching 1d index and data slices.
cupy.choose
cupy.diag	Returns a diagonal or a diagonal array.
cupy.diagonal	Returns specified diagonals.
cupy.lib.stride_tricks.as_strided	Create a view into the array with the given shape and strides.
cupy.place	Change elements of an array based on conditional and input values.
cupy.put	Replaces specified elements of an array with given values.
cupy.fill_diagonal	Fills the main diagonal of the given array of any dimensionality.

Input and Output

NumPy binary files (NPY, NPZ)

cupy.load	Loads arrays or pickled objects from .npy, .npz or pickled file.
cupy.save	Saves an array to a binary file in .npy format.
cupy.savez	Saves one or more arrays into a file in uncompressed .npz format.
cupy.savez_compressed	Saves one or more arrays into a file in compressed .npz format.

String formatting

cupy.array_repr	Returns the string representation of an array.
cupy.array_str	Returns the string representation of the content of an array.

Base-n representations

cupy.binary_repr	Return the binary representation of the input number as a string.
cupy.base_repr	Return a string representation of a number in the given base system.

Linear Algebra

Matrix and vector products

cupy.cross	Returns the cross product of two vectors.
cupy.dot	Returns a dot product of two arrays.
cupy.vdot	Returns the dot product of two vectors.
cupy.inner	Returns the inner product of two arrays.
cupy.outer	Returns the outer product of two vectors.
cupy.matmul	Returns the matrix product of two arrays and is the implementation of the @ operator introduced in Python 3.5 following PEP465.
cupy.tensordot	Returns the tensor dot product of two arrays along specified axes.
cupy.einsum	Evaluates the Einstein summation convention on the operands.
cupy.linalg.matrix_power	Raise a square matrix to the (integer) power n.
cupy.kron	Returns the kronecker product of two arrays.

Decompositions

cupy.linalg.cholesky	Cholesky decomposition.
cupy.linalg.qr	QR decomposition.
cupy.linalg.svd	Singular Value Decomposition.

Matrix eigenvalues

cupy.linalg.eigh	Eigenvalues and eigenvectors of a symmetric matrix.
cupy.linalg.eigvalsh	Calculates eigenvalues of a symmetric matrix.

Norms etc.

cupy.linalg.det	Returns the determinant of an array.
cupy.linalg.norm	Returns one of matrix norms specified by ord parameter.
cupy.linalg.matrix_rank	Return matrix rank of array using SVD method
cupy.linalg.slogdet	Returns sign and logarithm of the determinant of an array.
cupy.trace	Returns the sum along the diagonals of an array.

Solving linear equations

cupy.linalg.solve	Solves a linear matrix equation.
cupy.linalg.tensorsolve	Solves tensor equations denoted by ax = b.
cupy.linalg.lstsq	Return the least-squares solution to a linear matrix equation.
cupy.linalg.inv	Computes the inverse of a matrix.
cupy.linalg.pinv	Compute the Moore-Penrose pseudoinverse of a matrix.
cupy.linalg.tensorinv	Computes the inverse of a tensor.
cupyx.scipy.linalg.lu_factor	LU decomposition.
cupyx.scipy.linalg.lu_solve	Solve an equation system, a * x = b, given the LU factorization of a
cupyx.scipy.linalg.solve_triangular	Solve the equation a x = b for x, assuming a is a triangular matrix.

Logic Functions

Truth value testing

cupy.all	Tests whether all array elements along a given axis evaluate to True.
cupy.any	Tests whether any array elements along a given axis evaluate to True.
cupy.in1d	Tests whether each element of a 1-D array is also present in a second array.
cupy.isin	Calculates element in test_elements, broadcasting over element only.

Infinities and NaNs

cupy.isfinite	Tests finiteness elementwise.
cupy.isinf	Tests if each element is the positive or negative infinity.
cupy.isnan	Tests if each element is a NaN.

Array type testing

cupy.iscomplex	Returns a bool array, where True if input element is complex.
cupy.iscomplexobj	Check for a complex type or an array of complex numbers.
cupy.isfortran	Returns True if the array is Fortran contiguous but not C contiguous.
cupy.isreal	Returns a bool array, where True if input element is real.
cupy.isrealobj	Return True if x is a not complex type or an array of complex numbers.
cupy.isscalar	Returns True if the type of num is a scalar type.

Logic operations

cupy.logical_and	Computes the logical AND of two arrays.
cupy.logical_or	Computes the logical OR of two arrays.
cupy.logical_not	Computes the logical NOT of an array.
cupy.logical_xor	Computes the logical XOR of two arrays.

Comparison

cupy.allclose	Returns True if two arrays are element-wise equal within a tolerance.
cupy.isclose	Returns a boolean array where two arrays are equal within a tolerance.
cupy.greater	Tests elementwise if x1 > x2.
cupy.greater_equal	Tests elementwise if x1 >= x2.
cupy.less	Tests elementwise if x1 < x2.
cupy.less_equal	Tests elementwise if x1 <= x2.
cupy.equal	Tests elementwise if x1 == x2.
cupy.not_equal	Tests elementwise if x1 != x2.

Mathematical Functions

Trigonometric functions

cupy.sin	Elementwise sine function.
cupy.cos	Elementwise cosine function.
cupy.tan	Elementwise tangent function.
cupy.arcsin	Elementwise inverse-sine function (a.k.a.
cupy.arccos	Elementwise inverse-cosine function (a.k.a.
cupy.arctan	Elementwise inverse-tangent function (a.k.a.
cupy.hypot	Computes the hypoteneous of orthogonal vectors of given length.
cupy.arctan2	Elementwise inverse-tangent of the ratio of two arrays.
cupy.degrees	Converts angles from radians to degrees elementwise.
cupy.radians	Converts angles from degrees to radians elementwise.
cupy.unwrap	Unwrap by changing deltas between values to 2*pi complement.
cupy.deg2rad	Converts angles from degrees to radians elementwise.
cupy.rad2deg	Converts angles from radians to degrees elementwise.

Hyperbolic functions

cupy.sinh	Elementwise hyperbolic sine function.
cupy.cosh	Elementwise hyperbolic cosine function.
cupy.tanh	Elementwise hyperbolic tangent function.
cupy.arcsinh	Elementwise inverse of hyperbolic sine function.
cupy.arccosh	Elementwise inverse of hyperbolic cosine function.
cupy.arctanh	Elementwise inverse of hyperbolic tangent function.

Rounding

cupy.around	Rounds to the given number of decimals.
cupy.round_
cupy.rint	Rounds each element of an array to the nearest integer.
cupy.fix	If given value x is positive, it return floor(x).
cupy.floor	Rounds each element of an array to its floor integer.
cupy.ceil	Rounds each element of an array to its ceiling integer.
cupy.trunc	Rounds each element of an array towards zero.

Sums, products, differences

cupy.prod	Returns the product of an array along given axes.
cupy.sum	Returns the sum of an array along given axes.
cupy.cumprod	Returns the cumulative product of an array along a given axis.
cupy.cumsum	Returns the cumulative sum of an array along a given axis.
cupy.nansum	Returns the sum of an array along given axes treating Not a Numbers (NaNs) as zero.
cupy.nanprod	Returns the product of an array along given axes treating Not a Numbers (NaNs) as zero.
cupy.diff	Calculate the n-th discrete difference along the given axis.

Exponents and logarithms

cupy.exp	Elementwise exponential function.
cupy.expm1	Computes exp(x) - 1 elementwise.
cupy.exp2	Elementwise exponentiation with base 2.
cupy.log	Elementwise natural logarithm function.
cupy.log10	Elementwise common logarithm function.
cupy.log2	Elementwise binary logarithm function.
cupy.log1p	Computes log(1 + x) elementwise.
cupy.logaddexp	Computes log(exp(x1) + exp(x2)) elementwise.
cupy.logaddexp2	Computes log2(exp2(x1) + exp2(x2)) elementwise.

Other special functions

cupy.i0	Modified Bessel function of the first kind, order 0.
cupy.sinc	Elementwise sinc function.

Floating point routines

cupy.signbit	Tests elementwise if the sign bit is set (i.e.
cupy.copysign	Returns the first argument with the sign bit of the second elementwise.
cupy.frexp	Decomposes each element to mantissa and two’s exponent.
cupy.ldexp	Computes x1 * 2 ** x2 elementwise.
cupy.nextafter	Computes the nearest neighbor float values towards the second argument.

Arithmetic operations

cupy.add	Adds two arrays elementwise.
cupy.reciprocal	Computes 1 / x elementwise.
cupy.negative	Takes numerical negative elementwise.
cupy.multiply	Multiplies two arrays elementwise.
cupy.divide	Elementwise true division (i.e.
cupy.power	Computes x1 ** x2 elementwise.
cupy.subtract	Subtracts arguments elementwise.
cupy.true_divide	Elementwise true division (i.e.
cupy.floor_divide	Elementwise floor division (i.e.
cupy.fmod	Computes the remainder of C division elementwise.
cupy.mod	Computes the remainder of Python division elementwise.
cupy.modf	Extracts the fractional and integral parts of an array elementwise.
cupy.remainder	Computes the remainder of Python division elementwise.
cupy.divmod

Handling complex numbers

cupy.angle	Returns the angle of the complex argument.
cupy.real	Returns the real part of the elements of the array.
cupy.imag	Returns the imaginary part of the elements of the array.
cupy.conj	Returns the complex conjugate, element-wise.

Miscellaneous

cupy.clip	Clips the values of an array to a given interval.
cupy.sqrt	Elementwise square root function.
cupy.cbrt	Elementwise cube root function.
cupy.square	Elementwise square function.
cupy.absolute	Elementwise absolute value function.
cupy.sign	Elementwise sign function.
cupy.maximum	Takes the maximum of two arrays elementwise.
cupy.minimum	Takes the minimum of two arrays elementwise.
cupy.fmax	Takes the maximum of two arrays elementwise.
cupy.fmin	Takes the minimum of two arrays elementwise.
cupy.nan_to_num	Elementwise nan_to_num function.
cupy.blackman	Returns the Blackman window.
cupy.hamming	Returns the Hamming window.
cupy.hanning	Returns the Hanning window.

Padding

cupy.pad

Pads an array with specified widths and values.

Random Sampling (cupy.random)

Differences between cupy.random and numpy.random:

• CuPy provides Legacy Random Generation API (see also: NumPy 1.16 Reference). The new random generator API (numpy.random.Generator class) introduced in NumPy 1.17 has not been implemented yet.
• Most functions under cupy.random support the dtype option, which do not exist in the corresponding NumPy APIs. This option enables generation of float32 values directly without any space overhead.
• CuPy does not guarantee that the same number generator is used across major versions. This means that numbers generated by cupy.random by new major version may not be the same as the previous one, even if the same seed and distribution are used.

Simple random data

cupy.random.rand	Returns an array of uniform random values over the interval [0, 1).
cupy.random.randn	Returns an array of standard normal random values.
cupy.random.randint	Returns a scalar or an array of integer values over [low, high).
cupy.random.random_integers	Return a scalar or an array of integer values over [low, high]
cupy.random.random_sample	Returns an array of random values over the interval [0, 1).
cupy.random.random	Returns an array of random values over the interval [0, 1).
cupy.random.ranf	Returns an array of random values over the interval [0, 1).
cupy.random.sample	Returns an array of random values over the interval [0, 1).
cupy.random.choice	Returns an array of random values from a given 1-D array.
cupy.random.bytes	Returns random bytes.

Permutations

cupy.random.shuffle	Shuffles an array.
cupy.random.permutation	Returns a permuted range or a permutation of an array.

Distributions

cupy.random.beta	Beta distribution.
cupy.random.binomial	Binomial distribution.
cupy.random.chisquare	Chi-square distribution.
cupy.random.dirichlet	Dirichlet distribution.
cupy.random.exponential	Exponential distribution.
cupy.random.f	F distribution.
cupy.random.gamma	Gamma distribution.
cupy.random.geometric	Geometric distribution.
cupy.random.gumbel	Returns an array of samples drawn from a Gumbel distribution.
cupy.random.hypergeometric	hypergeometric distribution.
cupy.random.laplace	Laplace distribution.
cupy.random.logistic	Logistic distribution.
cupy.random.lognormal	Returns an array of samples drawn from a log normal distribution.
cupy.random.logseries	Log series distribution.
cupy.random.multinomial	Returns an array from multinomial distribution.
cupy.random.multivariate_normal	(experimental) Multivariate normal distribution.
cupy.random.negative_binomial	Negative binomial distribution.
cupy.random.noncentral_chisquare	Noncentral chisquare distribution.
cupy.random.noncentral_f	Noncentral F distribution.
cupy.random.normal	Returns an array of normally distributed samples.
cupy.random.pareto	Pareto II or Lomax distribution.
cupy.random.poisson	Poisson distribution.
cupy.random.power	Power distribution.
cupy.random.rayleigh	Rayleigh distribution.
cupy.random.standard_cauchy	Standard cauchy distribution.
cupy.random.standard_exponential	Standard exponential distribution.
cupy.random.standard_gamma	Standard gamma distribution.
cupy.random.standard_normal	Returns an array of samples drawn from the standard normal distribution.
cupy.random.standard_t	Standard Student’s t distribution.
cupy.random.triangular	Triangular distribution.
cupy.random.uniform	Returns an array of uniformly-distributed samples over an interval.
cupy.random.vonmises	von Mises distribution.
cupy.random.wald	Wald distribution.
cupy.random.weibull	weibull distribution.
cupy.random.zipf	Zipf distribution.

Random generator

cupy.random.RandomState	Portable container of a pseudo-random number generator.
cupy.random.seed	Resets the state of the random number generator with a seed.
cupy.random.get_random_state	Gets the state of the random number generator for the current device.
cupy.random.set_random_state	Sets the state of the random number generator for the current device.

Note

CuPy does not provide cupy.random.get_state nor cupy.random.set_state at this time. Use cupy.random.get_random_state() and cupy.random.set_random_state() instead. Note that these functions use cupy.random.RandomState instance to represent the internal state, which cannot be serialized.

Sorting, Searching, and Counting

Sorting

cupy.sort	Returns a sorted copy of an array with a stable sorting algorithm.
cupy.lexsort	Perform an indirect sort using an array of keys.
cupy.argsort	Returns the indices that would sort an array with a stable sorting.
cupy.msort	Returns a copy of an array sorted along the first axis.
cupy.partition	Returns a partitioned copy of an array.
cupy.argpartition	Returns the indices that would partially sort an array.

See also

cupy.ndarray.sort()

Searching

cupy.argmax	Returns the indices of the maximum along an axis.
cupy.nanargmax	Return the indices of the maximum values in the specified axis ignoring NaNs.
cupy.argmin	Returns the indices of the minimum along an axis.
cupy.nanargmin	Return the indices of the minimum values in the specified axis ignoring NaNs.
cupy.nonzero	Return the indices of the elements that are non-zero.
cupy.flatnonzero	Return indices that are non-zero in the flattened version of a.
cupy.where	Return elements, either from x or y, depending on condition.

Counting

cupy.count_nonzero

Counts the number of non-zero values in the array.

Statistical Functions

Order statistics

cupy.amin	Returns the minimum of an array or the minimum along an axis.
cupy.amax	Returns the maximum of an array or the maximum along an axis.
cupy.nanmin	Returns the minimum of an array along an axis ignoring NaN.
cupy.nanmax	Returns the maximum of an array along an axis ignoring NaN.
cupy.percentile	Computes the q-th percentile of the data along the specified axis.

Means and variances

cupy.average	Returns the weighted average along an axis.
cupy.mean	Returns the arithmetic mean along an axis.
cupy.var	Returns the variance along an axis.
cupy.std	Returns the standard deviation along an axis.
cupy.nanmean	Returns the arithmetic mean along an axis ignoring NaN values.
cupy.nanvar	Returns the variance along an axis ignoring NaN values.
cupy.nanstd	Returns the standard deviation along an axis ignoring NaN values.

Histograms

cupy.histogram	Computes the histogram of a set of data.
cupy.bincount	Count number of occurrences of each value in array of non-negative ints.

Correlations

cupy.corrcoef	Returns the Pearson product-moment correlation coefficients of an array.
cupy.cov	Returns the covariance matrix of an array.

CuPy-specific Functions

CuPy-specific functions are placed under cupyx namespace.

cupyx.rsqrt	Returns the reciprocal square root.
cupyx.scatter_add	Adds given values to specified elements of an array.
cupyx.scatter_max	Stores a maximum value of elements specified by indices to an array.
cupyx.scatter_min	Stores a minimum value of elements specified by indices to an array.

SciPy-compatible Routines

The following pages describe SciPy-compatible routines. These functions cover a subset of SciPy routines.

Discrete Fourier transforms (scipy.fft)

Fast Fourier Transforms

cupyx.scipy.fft.fft	Compute the one-dimensional FFT.
cupyx.scipy.fft.ifft	Compute the one-dimensional inverse FFT.
cupyx.scipy.fft.fft2	Compute the two-dimensional FFT.
cupyx.scipy.fft.ifft2	Compute the two-dimensional inverse FFT.
cupyx.scipy.fft.fftn	Compute the N-dimensional FFT.
cupyx.scipy.fft.ifftn	Compute the N-dimensional inverse FFT.
cupyx.scipy.fft.rfft	Compute the one-dimensional FFT for real input.
cupyx.scipy.fft.irfft	Compute the one-dimensional inverse FFT for real input.
cupyx.scipy.fft.rfft2	Compute the two-dimensional FFT for real input.
cupyx.scipy.fft.irfft2	Compute the two-dimensional inverse FFT for real input.
cupyx.scipy.fft.rfftn	Compute the N-dimensional FFT for real input.
cupyx.scipy.fft.irfftn	Compute the N-dimensional inverse FFT for real input.
cupyx.scipy.fft.hfft	Compute the FFT of a signal that has Hermitian symmetry.
cupyx.scipy.fft.ihfft	Compute the FFT of a signal that has Hermitian symmetry.

Code compatibility features

1. The boolean switch cupy.fft.config.enable_nd_planning also affects the FFT functions in this module, see FFT Functions. Moreover, as with other FFT modules in CuPy, FFT functions in this module can take advantage of an existing cuFFT plan (returned by cupyx.scipy.fftpack.get_fft_plan()) when used as a context manager.
2. Like in scipy.fft, all FFT functions in this module have an optional argument overwrite_x (default is False), which has the same semantics as in scipy.fft: when it is set to True, the input array x can (not will) be overwritten arbitrarily. This is not an in-place FFT, the user should always use the return value from the functions, e.g. x = cupyx.scipy.fft.fft(x, ..., overwrite_x=True, ...).
3. The cupyx.scipy.fft module can also be used as a backend for scipy.fft e.g. by installing with scipy.fft.set_backend(cupyx.scipy.fft). This can allow scipy.fft to work with both numpy and cupy arrays.

Note

scipy.fft requires SciPy version 1.4.0 or newer.

Legacy Discrete Fourier transforms (scipy.fftpack)

Note

As of SciPy version 1.4.0, scipy.fft is recommended over scipy.fftpack. Consider using cupyx.scipy.fft instead.

Fast Fourier Transforms

cupyx.scipy.fftpack.fft	Compute the one-dimensional FFT.
cupyx.scipy.fftpack.ifft	Compute the one-dimensional inverse FFT.
cupyx.scipy.fftpack.fft2	Compute the two-dimensional FFT.
cupyx.scipy.fftpack.ifft2	Compute the two-dimensional inverse FFT.
cupyx.scipy.fftpack.fftn	Compute the N-dimensional FFT.
cupyx.scipy.fftpack.ifftn	Compute the N-dimensional inverse FFT.
cupyx.scipy.fftpack.rfft	Compute the one-dimensional FFT for real input.
cupyx.scipy.fftpack.irfft	Compute the one-dimensional inverse FFT for real input.
cupyx.scipy.fftpack.get_fft_plan	Generate a CUDA FFT plan for transforming up to three axes.

Code compatibility features

1. The get_fft_plan function has no counterpart in scipy.fftpack. It returns a cuFFT plan that can be passed to the FFT functions in this module (using the argument plan) to accelarate the computation. The argument plan is currently experimental and the interface may be changed in the future version.
2. The boolean switch cupy.fft.config.enable_nd_planning also affects the FFT functions in this module, see FFT Functions. This switch is neglected when planning manually using get_fft_plan.
3. Like in scipy.fftpack, all FFT functions in this module have an optional argument overwrite_x (default is False), which has the same semantics as in scipy.fftpack: when it is set to True, the input array x can (not will) be destroyed and replaced by the output. For this reason, when an in-place FFT is desired, the user should always reassign the input in the following manner: x = cupyx.scipy.fftpack.fft(x, ..., overwrite_x=True, ...).

Multi-dimensional image processing

CuPy provides multi-dimensional image processing functions. It supports a subset of scipy.ndimage interface.

Interpolation

cupyx.scipy.ndimage.affine_transform	Apply an affine transformation.
cupyx.scipy.ndimage.convolve	Multi-dimensional convolution.
cupyx.scipy.ndimage.correlate	Multi-dimensional correlate.
cupyx.scipy.ndimage.map_coordinates	Map the input array to new coordinates by interpolation.
cupyx.scipy.ndimage.rotate	Rotate an array.
cupyx.scipy.ndimage.shift	Shift an array.
cupyx.scipy.ndimage.zoom	Zoom an array.

OpenCV mode

cupyx.scipy.ndimage supports additional mode, opencv. If it is given, the function performs like cv2.warpAffine or cv2.resize.

Sparse matrices

CuPy supports sparse matrices using cuSPARSE. These matrices have the same interfaces of SciPy’s sparse matrices.

Conversion to/from SciPy sparse matrices

cupyx.scipy.sparse.*_matrix and scipy.sparse.*_matrix are not implicitly convertible to each other. That means, SciPy functions cannot take cupyx.scipy.sparse.*_matrix objects as inputs, and vice versa.

• To convert SciPy sparse matrices to CuPy, pass it to the constructor of each CuPy sparse matrix class.
• To convert CuPy sparse matrices to SciPy, use get method of each CuPy sparse matrix class.

Note that converting between CuPy and SciPy incurs data transfer between the host (CPU) device and the GPU device, which is costly in terms of performance.

Conversion to/from CuPy ndarrays

• To convert CuPy ndarray to CuPy sparse matrices, pass it to the constructor of each CuPy sparse matrix class.
• To convert CuPy sparse matrices to CuPy ndarray, use toarray of each CuPy sparse matrix instance (e.g., cupyx.scipy.sparse.csr_matrix.toarray()).

Converting between CuPy ndarray and CuPy sparse matrices does not incur data transfer; it is copied inside the GPU device.

Sparse matrix classes

cupyx.scipy.sparse.coo_matrix	COOrdinate format sparse matrix.
cupyx.scipy.sparse.csc_matrix	Compressed Sparse Column matrix.
cupyx.scipy.sparse.csr_matrix	Compressed Sparse Row matrix.
cupyx.scipy.sparse.dia_matrix	Sparse matrix with DIAgonal storage.
cupyx.scipy.sparse.spmatrix	Base class of all sparse matrixes.

Functions

Building sparse matrices

cupyx.scipy.sparse.diags	Construct a sparse matrix from diagonals.
cupyx.scipy.sparse.eye	Creates a sparse matrix with ones on diagonal.
cupyx.scipy.sparse.identity	Creates an identity matrix in sparse format.
cupyx.scipy.sparse.spdiags	Creates a sparse matrix from diagonals.
cupyx.scipy.sparse.rand	Generates a random sparse matrix.
cupyx.scipy.sparse.random	Generates a random sparse matrix.

Identifying sparse matrices

cupyx.scipy.sparse.issparse	Checks if a given matrix is a sparse matrix.
cupyx.scipy.sparse.isspmatrix	Checks if a given matrix is a sparse matrix.
cupyx.scipy.sparse.isspmatrix_csc	Checks if a given matrix is of CSC format.
cupyx.scipy.sparse.isspmatrix_csr	Checks if a given matrix is of CSR format.
cupyx.scipy.sparse.isspmatrix_coo	Checks if a given matrix is of COO format.
cupyx.scipy.sparse.isspmatrix_dia	Checks if a given matrix is of DIA format.

Linear Algebra

cupyx.scipy.sparse.linalg.lsqr

Solves linear system with QR decomposition.

Special Functions

Bessel Functions

cupyx.scipy.special.j0	Bessel function of the first kind of order 0.
cupyx.scipy.special.j1	Bessel function of the first kind of order 1.
cupyx.scipy.special.y0	Bessel function of the second kind of order 0.
cupyx.scipy.special.y1	Bessel function of the second kind of order 1.
cupyx.scipy.special.i0	Modified Bessel function of order 0.
cupyx.scipy.special.i1	Modified Bessel function of order 1.

Gamma and Related Functions

cupyx.scipy.special.gamma	Gamma function.
cupyx.scipy.special.gammaln	Logarithm of the absolute value of the Gamma function.
cupyx.scipy.special.polygamma	Polygamma function n.
cupyx.scipy.special.digamma	The digamma function.

Raw Statistical Functions

cupyx.scipy.special.ndtr

Cumulative distribution function of normal distribution.

Error Function

cupyx.scipy.special.erf	Error function.
cupyx.scipy.special.erfc	Complementary error function.
cupyx.scipy.special.erfcx	Scaled complementary error function.
cupyx.scipy.special.erfinv	Inverse function of error function.
cupyx.scipy.special.erfcinv	Inverse function of complementary error function.

Other Special Functions

cupyx.scipy.special.zeta

Hurwitz zeta function.

NumPy-CuPy Generic Code Support

cupy.get_array_module	Returns the array module for arguments.
cupyx.scipy.get_array_module	Returns the array module for arguments.

Memory Management

CuPy uses memory pool for memory allocations by default. The memory pool significantly improves the performance by mitigating the overhead of memory allocation and CPU/GPU synchronization.

There are two different memory pools in CuPy:

• Device memory pool (GPU device memory), which is used for GPU memory allocations.
• Pinned memory pool (non-swappable CPU memory), which is used during CPU-to-GPU data transfer.

Attention

When you monitor the memory usage (e.g., using nvidia-smi for GPU memory or ps for CPU memory), you may notice that memory not being freed even after the array instance become out of scope. This is an expected behavior, as the default memory pool “caches” the allocated memory blocks.

See Low-Level CUDA Support for the details of memory management APIs.

Memory Pool Operations

The memory pool instance provides statistics about memory allocation. To access the default memory pool instance, use cupy.get_default_memory_pool() and cupy.get_default_pinned_memory_pool(). You can also free all unused memory blocks hold in the memory pool. See the example code below for details:

import cupy
import numpy

mempool = cupy.get_default_memory_pool()
pinned_mempool = cupy.get_default_pinned_memory_pool()

# Create an array on CPU.
# NumPy allocates 400 bytes in CPU (not managed by CuPy memory pool).
a_cpu = numpy.ndarray(100, dtype=numpy.float32)
print(a_cpu.nbytes)                      # 400

# You can access statistics of these memory pools.
print(mempool.used_bytes())              # 0
print(mempool.total_bytes())             # 0
print(pinned_mempool.n_free_blocks())    # 0

# Transfer the array from CPU to GPU.
# This allocates 400 bytes from the device memory pool, and another 400
# bytes from the pinned memory pool.  The allocated pinned memory will be
# released just after the transfer is complete.  Note that the actual
# allocation size may be rounded to larger value than the requested size
# for performance.
a = cupy.array(a_cpu)
print(a.nbytes)                          # 400
print(mempool.used_bytes())              # 512
print(mempool.total_bytes())             # 512
print(pinned_mempool.n_free_blocks())    # 1

# When the array goes out of scope, the allocated device memory is released
# and kept in the pool for future reuse.
a = None  # (or `del a`)
print(mempool.used_bytes())              # 0
print(mempool.total_bytes())             # 512
print(pinned_mempool.n_free_blocks())    # 1

# You can clear the memory pool by calling `free_all_blocks`.
mempool.free_all_blocks()
pinned_mempool.free_all_blocks()
print(mempool.used_bytes())              # 0
print(mempool.total_bytes())             # 0
print(pinned_mempool.n_free_blocks())    # 0

See cupy.cuda.MemoryPool and cupy.cuda.PinnedMemoryPool for details.

Limiting GPU Memory Usage

You can hard-limit the amount of GPU memory that can be allocated by using CUPY_GPU_MEMORY_LIMIT environment variable (see Environment variables for details).

# Set the hard-limit to 1 GiB:
#   $ export CUPY_GPU_MEMORY_LIMIT="1073741824"

# You can also specify the limit in fraction of the total amount of memory
# on the GPU. If you have a GPU with 2 GiB memory, the following is
# equivalent to the above configuration.
#   $ export CUPY_GPU_MEMORY_LIMIT="50%"

import cupy
print(cupy.get_default_memory_pool().get_limit())  # 1073741824

You can also set the limit (or override the value specified via the environment variable) using cupy.cuda.MemoryPool.set_limit(). In this way, you can use a different limit for each GPU device.

import cupy

mempool = cupy.get_default_memory_pool()

with cupy.cuda.Device(0):
    mempool.set_limit(size=1024**3)  # 1 GiB

with cupy.cuda.Device(1):
    mempool.set_limit(size=2*1024**3)  # 2 GiB

Note

CUDA allocates some GPU memory outside of the memory pool (such as CUDA context, library handles, etc.). Depending on the usage, such memory may take one to few hundred MiB. That will not be counted in the limit.

Changing Memory Pool

You can use your own memory allocator instead of the default memory pool by passing the memory allocation function to cupy.cuda.set_allocator() / cupy.cuda.set_pinned_memory_allocator(). The memory allocator function should take 1 argument (the requested size in bytes) and return cupy.cuda.MemoryPointer / cupy.cuda.PinnedMemoryPointer.

You can even disable the default memory pool by the code below. Be sure to do this before any other CuPy operations.

import cupy

# Disable memory pool for device memory (GPU)
cupy.cuda.set_allocator(None)

# Disable memory pool for pinned memory (CPU).
cupy.cuda.set_pinned_memory_allocator(None)

Low-Level CUDA Support

Device management

cupy.cuda.Device

Object that represents a CUDA device.

Memory management

cupy.get_default_memory_pool	Returns CuPy default memory pool for GPU memory.
cupy.get_default_pinned_memory_pool	Returns CuPy default memory pool for pinned memory.
cupy.cuda.Memory	Memory allocation on a CUDA device.
cupy.cuda.UnownedMemory	CUDA memory that is not owned by CuPy.
cupy.cuda.PinnedMemory	Pinned memory allocation on host.
cupy.cuda.MemoryPointer	Pointer to a point on a device memory.
cupy.cuda.PinnedMemoryPointer	Pointer of a pinned memory.
cupy.cuda.alloc	Calls the current allocator.
cupy.cuda.alloc_pinned_memory	Calls the current allocator.
cupy.cuda.get_allocator	Returns the current allocator for GPU memory.
cupy.cuda.set_allocator	Sets the current allocator for GPU memory.
cupy.cuda.using_allocator	Sets a thread-local allocator for GPU memory inside
cupy.cuda.set_pinned_memory_allocator	Sets the current allocator for the pinned memory.
cupy.cuda.MemoryPool	Memory pool for all GPU devices on the host.
cupy.cuda.PinnedMemoryPool	Memory pool for pinned memory on the host.

Memory hook

cupy.cuda.MemoryHook	Base class of hooks for Memory allocations.
cupy.cuda.memory_hooks.DebugPrintHook	Memory hook that prints debug information.
cupy.cuda.memory_hooks.LineProfileHook	Code line CuPy memory profiler.

Streams and events

cupy.cuda.Stream	CUDA stream.
cupy.cuda.get_current_stream	Gets current CUDA stream.
cupy.cuda.Event	CUDA event, a synchronization point of CUDA streams.
cupy.cuda.get_elapsed_time	Gets the elapsed time between two events.

Texture memory

cupy.cuda.texture.ChannelFormatDescriptor	A class that holds the channel format description.
cupy.cuda.texture.CUDAarray	Allocate a CUDA array (cudaArray_t) that can be used as texture memory.
cupy.cuda.texture.ResourceDescriptor	A class that holds the resource description.
cupy.cuda.texture.TextureDescriptor	A class that holds the texture description.
cupy.cuda.texture.TextureObject	A class that holds a texture object.
cupy.cuda.texture.TextureReference	A class that holds a texture reference.

Profiler

cupy.cuda.profile	Enable CUDA profiling during with statement.
cupy.cuda.profiler.initialize	Initialize the CUDA profiler.
cupy.cuda.profiler.start	Enable profiling.
cupy.cuda.profiler.stop	Disable profiling.
cupy.cuda.nvtx.Mark	Marks an instantaneous event (marker) in the application.
cupy.cuda.nvtx.MarkC	Marks an instantaneous event (marker) in the application.
cupy.cuda.nvtx.RangePush	Starts a nested range.
cupy.cuda.nvtx.RangePushC	Starts a nested range.
cupy.cuda.nvtx.RangePop	Ends a nested range.

NCCL

cupy.cuda.nccl.NcclCommunicator	Initialize an NCCL communicator for one device controlled by one process.
cupy.cuda.nccl.get_build_version
cupy.cuda.nccl.get_version	Returns the runtime version of NCCL.
cupy.cuda.nccl.get_unique_id
cupy.cuda.nccl.groupStart	Start a group of NCCL calls.
cupy.cuda.nccl.groupEnd	End a group of NCCL calls.

Runtime API

CuPy wraps CUDA Runtime APIs to provide the native CUDA operations. Please check the Original CUDA Runtime API document to use these functions.

cupy.cuda.runtime.driverGetVersion
cupy.cuda.runtime.runtimeGetVersion
cupy.cuda.runtime.getDevice
cupy.cuda.runtime.deviceGetAttribute
cupy.cuda.runtime.deviceGetByPCIBusId
cupy.cuda.runtime.deviceGetPCIBusId
cupy.cuda.runtime.getDeviceCount
cupy.cuda.runtime.setDevice
cupy.cuda.runtime.deviceSynchronize
cupy.cuda.runtime.deviceCanAccessPeer
cupy.cuda.runtime.deviceEnablePeerAccess
cupy.cuda.runtime.malloc
cupy.cuda.runtime.mallocManaged
cupy.cuda.runtime.malloc3DArray
cupy.cuda.runtime.mallocArray
cupy.cuda.runtime.hostAlloc
cupy.cuda.runtime.hostRegister
cupy.cuda.runtime.hostUnregister
cupy.cuda.runtime.free
cupy.cuda.runtime.freeHost
cupy.cuda.runtime.freeArray
cupy.cuda.runtime.memGetInfo
cupy.cuda.runtime.memcpy
cupy.cuda.runtime.memcpyAsync
cupy.cuda.runtime.memcpyPeer
cupy.cuda.runtime.memcpyPeerAsync
cupy.cuda.runtime.memcpy2D
cupy.cuda.runtime.memcpy2DAsync
cupy.cuda.runtime.memcpy2DFromArray
cupy.cuda.runtime.memcpy2DFromArrayAsync
cupy.cuda.runtime.memcpy2DToArray
cupy.cuda.runtime.memcpy2DToArrayAsync
cupy.cuda.runtime.memcpy3D
cupy.cuda.runtime.memcpy3DAsync
cupy.cuda.runtime.memset
cupy.cuda.runtime.memsetAsync
cupy.cuda.runtime.memPrefetchAsync
cupy.cuda.runtime.memAdvise
cupy.cuda.runtime.pointerGetAttributes
cupy.cuda.runtime.streamCreate
cupy.cuda.runtime.streamCreateWithFlags
cupy.cuda.runtime.streamDestroy
cupy.cuda.runtime.streamSynchronize
cupy.cuda.runtime.streamAddCallback
cupy.cuda.runtime.streamQuery
cupy.cuda.runtime.streamWaitEvent
cupy.cuda.runtime.eventCreate
cupy.cuda.runtime.eventCreateWithFlags
cupy.cuda.runtime.eventDestroy
cupy.cuda.runtime.eventElapsedTime
cupy.cuda.runtime.eventQuery
cupy.cuda.runtime.eventRecord
cupy.cuda.runtime.eventSynchronize

Kernel binary memoization

cupy.memoize	Makes a function memoizing the result for each argument and device.
cupy.clear_memo	Clears the memoized results for all functions decorated by memoize.

Custom kernels

cupy.ElementwiseKernel	User-defined elementwise kernel.
cupy.ReductionKernel	User-defined reduction kernel.
cupy.RawKernel	User-defined custom kernel.
cupy.RawModule	User-defined custom module.
cupy.fuse	Decorator that fuses a function.

Interoperability

CuPy can also be used in conjunction with other frameworks.

NumPy

cupy.ndarray implements __array_ufunc__ interface (see NEP 13 — A Mechanism for Overriding Ufuncs for details). This enables NumPy ufuncs to be directly operated on CuPy arrays. __array_ufunc__ feature requires NumPy 1.13 or later.

import cupy
import numpy

arr = cupy.random.randn(1, 2, 3, 4).astype(cupy.float32)
result = numpy.sum(arr)
print(type(result))  # => <class 'cupy.core.core.ndarray'>

cupy.ndarray also implements __array_function__ interface (see NEP 18 — A dispatch mechanism for NumPy’s high level array functions for details). This enables code using NumPy to be directly operated on CuPy arrays. __array_function__ feature requires NumPy 1.16 or later; note that this is currently defined as an experimental feature of NumPy and you need to specify the environment variable (NUMPY_EXPERIMENTAL_ARRAY_FUNCTION=1) to enable it.

Numba

Numba is a Python JIT compiler with NumPy support.

cupy.ndarray implements __cuda_array_interface__, which is the CUDA array interchange interface compatible with Numba v0.39.0 or later (see CUDA Array Interface for details). It means you can pass CuPy arrays to kernels JITed with Numba. The following is a simple example code borrowed from numba/numba#2860:

import cupy
from numba import cuda

@cuda.jit
def add(x, y, out):
        start = cuda.grid(1)
        stride = cuda.gridsize(1)
        for i in range(start, x.shape[0], stride):
                out[i] = x[i] + y[i]

a = cupy.arange(10)
b = a * 2
out = cupy.zeros_like(a)

print(out)  # => [0 0 0 0 0 0 0 0 0 0]

add[1, 32](a, b, out)

print(out)  # => [ 0  3  6  9 12 15 18 21 24 27]

In addition, cupy.asarray() supports zero-copy conversion from Numba CUDA array to CuPy array.

import numpy
import numba
import cupy

x = numpy.arange(10)  # type: numpy.ndarray
x_numba = numba.cuda.to_device(x)  # type: numba.cuda.cudadrv.devicearray.DeviceNDArray
x_cupy = cupy.asarray(x_numba)  # type: cupy.ndarray

mpi4py

MPI for Python (mpi4py) is a Python wrapper for the Message Passing Interface (MPI) libraries.

MPI is the most widely used standard for high-performance inter-process communications. Recently several MPI vendors, including Open MPI and MVAPICH, have extended their support beyond the v3.1 standard to enable “CUDA-awareness”; that is, passing CUDA device pointers directly to MPI calls to avoid explicit data movement between the host and the device.

With the aforementioned __cuda_array_interface__ standard implemented in CuPy, mpi4py now provides (experimental) support for passing CuPy arrays to MPI calls, provided that mpi4py is built against a CUDA-aware MPI implementation. The folowing is a simple example code borrowed from mpi4py Tutorial:

# To run this script with N MPI processes, do
# mpiexec -n N python this_script.py

import cupy
from mpi4py import MPI

comm = MPI.COMM_WORLD
size = comm.Get_size()

# Allreduce
sendbuf = cupy.arange(10, dtype='i')
recvbuf = cupy.empty_like(sendbuf)
comm.Allreduce(sendbuf, recvbuf)
assert cupy.allclose(recvbuf, sendbuf*size)

This new feature will be officially released in mpi4py 3.1.0. To try it out, please build mpi4py from source for the time being. See the mpi4py website for more information.

DLPack

DLPack is a specification of tensor structure to share tensors among frameworks.

CuPy supports importing from and exporting to DLPack data structure (cupy.fromDlpack() and cupy.ndarray.toDlpack()).

cupy.fromDlpack

Zero-copy conversion from a DLPack tensor to a ndarray.

Here is a simple example:

import cupy

# Create a CuPy array.
cx1 = cupy.random.randn(1, 2, 3, 4).astype(cupy.float32)

# Convert it into a DLPack tensor.
dx = cx1.toDlpack()

# Convert it back to a CuPy array.
cx2 = cupy.fromDlpack(dx)

Here is an example of converting PyTorch tensor into cupy.ndarray.

import cupy
import torch

from torch.utils.dlpack import to_dlpack
from torch.utils.dlpack import from_dlpack

# Create a PyTorch tensor.
tx1 = torch.randn(1, 2, 3, 4).cuda()

# Convert it into a DLPack tensor.
dx = to_dlpack(tx1)

# Convert it into a CuPy array.
cx = cupy.fromDlpack(dx)

# Convert it back to a PyTorch tensor.
tx2 = from_dlpack(cx.toDlpack())

Testing Modules

CuPy offers testing utilities to support unit testing. They are under namespace cupy.testing.

Standard Assertions

The assertions have same names as NumPy’s ones. The difference from NumPy is that they can accept both numpy.ndarray and cupy.ndarray.

cupy.testing.assert_allclose	Raises an AssertionError if objects are not equal up to desired tolerance.
cupy.testing.assert_array_almost_equal	Raises an AssertionError if objects are not equal up to desired precision.
cupy.testing.assert_array_almost_equal_nulp	Compare two arrays relatively to their spacing.
cupy.testing.assert_array_max_ulp	Check that all items of arrays differ in at most N Units in the Last Place.
cupy.testing.assert_array_equal	Raises an AssertionError if two array_like objects are not equal.
cupy.testing.assert_array_list_equal	Compares lists of arrays pairwise with assert_array_equal.
cupy.testing.assert_array_less	Raises an AssertionError if array_like objects are not ordered by less than.

NumPy-CuPy Consistency Check

The following decorators are for testing consistency between CuPy’s functions and corresponding NumPy’s ones.

cupy.testing.numpy_cupy_allclose	Decorator that checks NumPy results and CuPy ones are close.
cupy.testing.numpy_cupy_array_almost_equal	Decorator that checks NumPy results and CuPy ones are almost equal.
cupy.testing.numpy_cupy_array_almost_equal_nulp	Decorator that checks results of NumPy and CuPy are equal w.r.t.
cupy.testing.numpy_cupy_array_max_ulp	Decorator that checks results of NumPy and CuPy ones are equal w.r.t.
cupy.testing.numpy_cupy_array_equal	Decorator that checks NumPy results and CuPy ones are equal.
cupy.testing.numpy_cupy_array_list_equal	Decorator that checks the resulting lists of NumPy and CuPy’s one are equal.
cupy.testing.numpy_cupy_array_less	Decorator that checks the CuPy result is less than NumPy result.
cupy.testing.numpy_cupy_raises	Decorator that checks the NumPy and CuPy throw same errors.

Parameterized dtype Test

The following decorators offer the standard way for parameterized test with respect to single or the combination of dtype(s).

cupy.testing.for_dtypes	Decorator for parameterized dtype test.
cupy.testing.for_all_dtypes	Decorator that checks the fixture with all dtypes.
cupy.testing.for_float_dtypes	Decorator that checks the fixture with float dtypes.
cupy.testing.for_signed_dtypes	Decorator that checks the fixture with signed dtypes.
cupy.testing.for_unsigned_dtypes	Decorator that checks the fixture with unsinged dtypes.
cupy.testing.for_int_dtypes	Decorator that checks the fixture with integer and optionally bool dtypes.
cupy.testing.for_complex_dtypes	Decorator that checks the fixture with complex dtypes.
cupy.testing.for_dtypes_combination	Decorator that checks the fixture with a product set of dtypes.
cupy.testing.for_all_dtypes_combination	Decorator that checks the fixture with a product set of all dtypes.
cupy.testing.for_signed_dtypes_combination	Decorator for parameterized test w.r.t.
cupy.testing.for_unsigned_dtypes_combination	Decorator for parameterized test w.r.t.
cupy.testing.for_int_dtypes_combination	Decorator for parameterized test w.r.t.

Parameterized order Test

The following decorators offer the standard way to parameterize tests with orders.

cupy.testing.for_orders	Decorator to parameterize tests with order.
cupy.testing.for_CF_orders	Decorator that checks the fixture with orders ‘C’ and ‘F’.

Profiling

time range

cupy.prof.TimeRangeDecorator	Decorator to mark function calls with range in NVIDIA profiler
cupy.prof.time_range	A context manager to describe the enclosed block as a nested range

Environment variables

Here are the environment variables CuPy uses.

CUDA_PATH	Path to the directory containing CUDA. The parent of the directory containing nvcc is used as default. When nvcc is not found, /usr/local/cuda is used. See Working with Custom CUDA Installation for details.
CUPY_CACHE_DIR	Path to the directory to store kernel cache. ${HOME}/.cupy/kernel_cache is used by default. See Overview for details.
CUPY_CACHE_SAVE_CUDA_SOURCE	If set to 1, CUDA source file will be saved along with compiled binary in the cache directory for debug purpose. It is disabled by default. Note: source file will not be saved if the compiled binary is already stored in the cache.
CUPY_DUMP_CUDA_SOURCE_ON_ERROR	If set to 1, when CUDA kernel compilation fails, CuPy dumps CUDA kernel code to standard error. It is disabled by default.
CUPY_CUDA_COMPILE_WITH_DEBUG	If set to 1, CUDA kernel will be compiled with debug information (--device-debug and --generate-line-info). It is disabled by default.
CUPY_GPU_MEMORY_LIMIT	The amount of memory that can be allocated for each device. The value can be specified in absolute bytes or fraction (e.g., "90%") of the total memory of each GPU. See Memory Management for details. 0 (unlimited) is used by default.
CUPY_SEED	Set the seed for random number generators. For historical reasons CHAINER_SEED is used if CUPY_SEED is unspecified.
CUPY_EXPERIMENTAL_SLICE_COPY	If set to 1, the following syntax is enabled: cupy_ndarray[:] = numpy_ndarray.

Moreover, as in any CUDA programs, all of the CUDA environment variables listed in the CUDA Toolkit Documentation will also be honored.

For installation

These environment variables are used during installation (building CuPy from source).

CUDA_PATH	See the description above.
CUTENSOR_PATH	Path to the cuTENSOR root directory that contains lib and include directories. (experimental)
NVCC	Define the compiler to use when compiling CUDA files.
CUPY_PYTHON_350_FORCE	Enforce CuPy to be installed against Python 3.5.0 (not recommended).
CUPY_INSTALL_USE_HIP	For building the ROCm support, see Install CuPy from Source for further detail.
CUPY_NVCC_GENERATE_CODE	To build CuPy for a particular CUDA architecture. For example, CUPY_NVCC_GENERATE_CODE=compute_60,sm_60. When this is not set, the default is to support all architectures.

Difference between CuPy and NumPy

The interface of CuPy is designed to obey that of NumPy. However, there are some differences.

Cast behavior from float to integer

Some casting behaviors from float to integer are not defined in C++ specification. The casting from a negative float to unsigned integer and infinity to integer is one of such examples. The behavior of NumPy depends on your CPU architecture. This is the result on an Intel CPU:

>>> np.array([-1], dtype=np.float32).astype(np.uint32)
array([4294967295], dtype=uint32)
>>> cupy.array([-1], dtype=np.float32).astype(np.uint32)
array([0], dtype=uint32)

>>> np.array([float('inf')], dtype=np.float32).astype(np.int32)
array([-2147483648], dtype=int32)
>>> cupy.array([float('inf')], dtype=np.float32).astype(np.int32)
array([2147483647], dtype=int32)

Random methods support dtype argument

NumPy’s random value generator does not support a dtype argument and instead always returns a float64 value. We support the option in CuPy because cuRAND, which is used in CuPy, supports both float32 and float64.

>>> np.random.randn(dtype=np.float32)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: randn() got an unexpected keyword argument 'dtype'
>>> cupy.random.randn(dtype=np.float32)    # doctest: +SKIP
array(0.10689262300729752, dtype=float32)

Out-of-bounds indices

CuPy handles out-of-bounds indices differently by default from NumPy when using integer array indexing. NumPy handles them by raising an error, but CuPy wraps around them.

>>> x = np.array([0, 1, 2])
>>> x[[1, 3]] = 10
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
IndexError: index 3 is out of bounds for axis 1 with size 3
>>> x = cupy.array([0, 1, 2])
>>> x[[1, 3]] = 10
>>> x
array([10, 10,  2])

Duplicate values in indices

CuPy’s __setitem__ behaves differently from NumPy when integer arrays reference the same location multiple times. In that case, the value that is actually stored is undefined. Here is an example of CuPy.

>>> a = cupy.zeros((2,))
>>> i = cupy.arange(10000) % 2
>>> v = cupy.arange(10000).astype(np.float32)
>>> a[i] = v
>>> a  # doctest: +SKIP
array([ 9150.,  9151.])

NumPy stores the value corresponding to the last element among elements referencing duplicate locations.

>>> a_cpu = np.zeros((2,))
>>> i_cpu = np.arange(10000) % 2
>>> v_cpu = np.arange(10000).astype(np.float32)
>>> a_cpu[i_cpu] = v_cpu
>>> a_cpu
array([9998., 9999.])

Zero-dimensional array

Reduction methods

NumPy’s reduction functions (e.g. numpy.sum()) return scalar values (e.g. numpy.float32). However CuPy counterparts return zero-dimensional cupy.ndarray s. That is because CuPy scalar values (e.g. cupy.float32) are aliases of NumPy scalar values and are allocated in CPU memory. If these types were returned, it would be required to synchronize between GPU and CPU. If you want to use scalar values, cast the returned arrays explicitly.

>>> type(np.sum(np.arange(3))) == np.int64
True
>>> type(cupy.sum(cupy.arange(3))) == cupy.core.core.ndarray
True

Type promotion

CuPy automatically promotes dtypes of cupy.ndarray s in a function with two or more operands, the result dtype is determined by the dtypes of the inputs. This is different from NumPy’s rule on type promotion, when operands contain zero-dimensional arrays. Zero-dimensional numpy.ndarray s are treated as if they were scalar values if they appear in operands of NumPy’s function, This may affect the dtype of its output, depending on the values of the “scalar” inputs.

>>> (np.array(3, dtype=np.int32) * np.array([1., 2.], dtype=np.float32)).dtype
dtype('float32')
>>> (np.array(300000, dtype=np.int32) * np.array([1., 2.], dtype=np.float32)).dtype
dtype('float64')
>>> (cupy.array(3, dtype=np.int32) * cupy.array([1., 2.], dtype=np.float32)).dtype
dtype('float64')

Data types

Data type of CuPy arrays cannot be non-numeric like strings and objects. See Overview for details.

Universal Functions only work with CuPy array or scalar

Unlike NumPy, Universal Functions in CuPy only work with CuPy array or scalar. They do not accept other objects (e.g., lists or numpy.ndarray).

>>> np.power([np.arange(5)], 2)
array([[ 0,  1,  4,  9, 16]])

>>> cupy.power([cupy.arange(5)], 2)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: Unsupported type <class 'list'>

Random seed arrays are hashed to scalars

Like Numpy, CuPy’s RandomState objects accept seeds either as numbers or as full numpy arrays.

>>> seed = np.array([1, 2, 3, 4, 5])
>>> rs = cupy.random.RandomState(seed=seed)

However, unlike Numpy, array seeds will be hashed down to a single number and so may not communicate as much entropy to the underlying random number generator.

Comparison Table

Here is a list of NumPy / SciPy APIs and its corresponding CuPy implementations.

- in CuPy column denotes that CuPy implementation is not provided yet. We welcome contributions for these functions.

NumPy / CuPy APIs

Module-Level

NumPy	CuPy
numpy.abs	cupy.abs
numpy.absolute	cupy.absolute
numpy.add	cupy.add
numpy.add_docstring	-
numpy.add_newdoc	-
numpy.add_newdoc_ufunc	-
numpy.alen	-
numpy.all	cupy.all
numpy.allclose	cupy.allclose
numpy.alltrue	-
numpy.amax	cupy.amax
numpy.amin	cupy.amin
numpy.angle	cupy.angle
numpy.any	cupy.any
numpy.append	-
numpy.apply_along_axis	-
numpy.apply_over_axes	-
numpy.arange	cupy.arange
numpy.arccos	cupy.arccos
numpy.arccosh	cupy.arccosh
numpy.arcsin	cupy.arcsin
numpy.arcsinh	cupy.arcsinh
numpy.arctan	cupy.arctan
numpy.arctan2	cupy.arctan2
numpy.arctanh	cupy.arctanh
numpy.argmax	cupy.argmax
numpy.argmin	cupy.argmin
numpy.argpartition	cupy.argpartition
numpy.argsort	cupy.argsort
numpy.argwhere	-
numpy.around	cupy.around
numpy.array	cupy.array
numpy.array2string	-
numpy.array_equal	-
numpy.array_equiv	-
numpy.array_repr	cupy.array_repr
numpy.array_split	cupy.array_split
numpy.array_str	cupy.array_str
numpy.asanyarray	cupy.asanyarray
numpy.asarray	cupy.asarray
numpy.asarray_chkfinite	-
numpy.ascontiguousarray	cupy.ascontiguousarray
numpy.asfarray	-
numpy.asfortranarray	cupy.asfortranarray
numpy.asmatrix	-
numpy.asscalar	-
numpy.atleast_1d	cupy.atleast_1d
numpy.atleast_2d	cupy.atleast_2d
numpy.atleast_3d	cupy.atleast_3d
numpy.average	cupy.average
numpy.bartlett	-
numpy.base_repr	cupy.base_repr
numpy.binary_repr	cupy.binary_repr
numpy.bincount	cupy.bincount
numpy.bitwise_and	cupy.bitwise_and
numpy.bitwise_not	-
numpy.bitwise_or	cupy.bitwise_or
numpy.bitwise_xor	cupy.bitwise_xor
numpy.blackman	cupy.blackman
numpy.block	-
numpy.bmat	-
numpy.broadcast_arrays	cupy.broadcast_arrays
numpy.broadcast_to	cupy.broadcast_to
numpy.busday_count	-
numpy.busday_offset	-
numpy.byte_bounds	-
numpy.can_cast	cupy.can_cast
numpy.cbrt	cupy.cbrt
numpy.ceil	cupy.ceil
numpy.choose	cupy.choose
numpy.clip	cupy.clip
numpy.column_stack	cupy.column_stack
numpy.common_type	cupy.common_type
numpy.compare_chararrays	-
numpy.compress	-
numpy.concatenate	cupy.concatenate
numpy.conj	cupy.conj
numpy.conjugate	-
numpy.convolve	-
numpy.copy	cupy.copy
numpy.copysign	cupy.copysign
numpy.copyto	cupy.copyto
numpy.corrcoef	cupy.corrcoef
numpy.correlate	-
numpy.cos	cupy.cos
numpy.cosh	cupy.cosh
numpy.count_nonzero	cupy.count_nonzero
numpy.cov	cupy.cov
numpy.cross	cupy.cross
numpy.cumprod	cupy.cumprod
numpy.cumproduct	-
numpy.cumsum	cupy.cumsum
numpy.datetime_as_string	-
numpy.datetime_data	-
numpy.deg2rad	cupy.deg2rad
numpy.degrees	cupy.degrees
numpy.delete	-
numpy.deprecate	-
numpy.deprecate_with_doc	-
numpy.diag	cupy.diag
numpy.diag_indices	-
numpy.diag_indices_from	-
numpy.diagflat	cupy.diagflat
numpy.diagonal	cupy.diagonal
numpy.diff	cupy.diff
numpy.digitize	-
numpy.disp	-
numpy.divide	cupy.divide
numpy.divmod	cupy.divmod
numpy.dot	cupy.dot
numpy.dsplit	cupy.dsplit
numpy.dstack	cupy.dstack
numpy.ediff1d	-
numpy.einsum	cupy.einsum
numpy.einsum_path	-
numpy.empty	cupy.empty
numpy.empty_like	cupy.empty_like
numpy.equal	cupy.equal
numpy.exp	cupy.exp
numpy.exp2	cupy.exp2
numpy.expand_dims	cupy.expand_dims
numpy.expm1	cupy.expm1
numpy.extract	-
numpy.eye	cupy.eye
numpy.fabs	-
numpy.fastCopyAndTranspose	-
numpy.fill_diagonal	cupy.fill_diagonal
numpy.find_common_type	cupy.find_common_type (alias of numpy.find_common_type)
numpy.fix	cupy.fix
numpy.flatnonzero	cupy.flatnonzero
numpy.flip	cupy.flip
numpy.fliplr	cupy.fliplr
numpy.flipud	cupy.flipud
numpy.float_power	-
numpy.floor	cupy.floor
numpy.floor_divide	cupy.floor_divide
numpy.fmax	cupy.fmax
numpy.fmin	cupy.fmin
numpy.fmod	cupy.fmod
numpy.format_float_positional	-
numpy.format_float_scientific	-
numpy.frexp	cupy.frexp
numpy.frombuffer	-
numpy.fromfile	cupy.fromfile
numpy.fromfunction	-
numpy.fromiter	-
numpy.frompyfunc	-
numpy.fromregex	-
numpy.fromstring	-
numpy.full	cupy.full
numpy.full_like	cupy.full_like
numpy.fv	-
numpy.gcd	-
numpy.genfromtxt	-
numpy.geomspace	-
numpy.get_array_wrap	-
numpy.get_include	-
numpy.get_printoptions	-
numpy.getbufsize	-
numpy.geterr	-
numpy.geterrcall	-
numpy.geterrobj	-
numpy.gradient	-
numpy.greater	cupy.greater
numpy.greater_equal	cupy.greater_equal
numpy.hamming	cupy.hamming
numpy.hanning	cupy.hanning
numpy.heaviside	-
numpy.histogram	cupy.histogram
numpy.histogram2d	-
numpy.histogram_bin_edges	-
numpy.histogramdd	-
numpy.hsplit	cupy.hsplit
numpy.hstack	cupy.hstack
numpy.hypot	cupy.hypot
numpy.i0	cupy.i0
numpy.identity	cupy.identity
numpy.imag	cupy.imag
numpy.in1d	cupy.in1d
numpy.indices	cupy.indices
numpy.info	-
numpy.inner	cupy.inner
numpy.insert	-
numpy.int_asbuffer	-
numpy.interp	-
numpy.intersect1d	-
numpy.invert	cupy.invert
numpy.ipmt	-
numpy.irr	-
numpy.is_busday	-
numpy.isclose	cupy.isclose
numpy.iscomplex	cupy.iscomplex
numpy.iscomplexobj	cupy.iscomplexobj
numpy.isfinite	cupy.isfinite
numpy.isfortran	cupy.isfortran
numpy.isin	cupy.isin
numpy.isinf	cupy.isinf
numpy.isnan	cupy.isnan
numpy.isnat	-
numpy.isneginf	-
numpy.isposinf	-
numpy.isreal	cupy.isreal
numpy.isrealobj	cupy.isrealobj
numpy.isscalar	cupy.isscalar
numpy.issctype	cupy.issctype (alias of numpy.issctype)
numpy.issubclass_	cupy.issubclass_ (alias of numpy.issubclass_)
numpy.issubdtype	cupy.issubdtype (alias of numpy.issubdtype)
numpy.issubsctype	cupy.issubsctype (alias of numpy.issubsctype)
numpy.iterable	-
numpy.ix_	cupy.ix_
numpy.kaiser	-
numpy.kron	cupy.kron
numpy.lcm	-
numpy.ldexp	cupy.ldexp
numpy.left_shift	cupy.left_shift
numpy.less	cupy.less
numpy.less_equal	cupy.less_equal
numpy.lexsort	cupy.lexsort
numpy.linspace	cupy.linspace
numpy.load	cupy.load
numpy.loads	-
numpy.loadtxt	-
numpy.log	cupy.log
numpy.log10	cupy.log10
numpy.log1p	cupy.log1p
numpy.log2	cupy.log2
numpy.logaddexp	cupy.logaddexp
numpy.logaddexp2	cupy.logaddexp2
numpy.logical_and	cupy.logical_and
numpy.logical_not	cupy.logical_not
numpy.logical_or	cupy.logical_or
numpy.logical_xor	cupy.logical_xor
numpy.logspace	cupy.logspace
numpy.lookfor	-
numpy.mafromtxt	-
numpy.mask_indices	-
numpy.mat	-
numpy.matmul	cupy.matmul
numpy.max	cupy.max
numpy.maximum	cupy.maximum
numpy.maximum_sctype	-
numpy.may_share_memory	cupy.may_share_memory
numpy.mean	cupy.mean
numpy.median	-
numpy.meshgrid	cupy.meshgrid
numpy.min	cupy.min
numpy.min_scalar_type	cupy.min_scalar_type (alias of numpy.min_scalar_type)
numpy.minimum	cupy.minimum
numpy.mintypecode	cupy.mintypecode (alias of numpy.mintypecode)
numpy.mirr	-
numpy.mod	cupy.mod
numpy.modf	cupy.modf
numpy.moveaxis	cupy.moveaxis
numpy.msort	cupy.msort
numpy.multiply	cupy.multiply
numpy.nan_to_num	cupy.nan_to_num
numpy.nanargmax	cupy.nanargmax
numpy.nanargmin	cupy.nanargmin
numpy.nancumprod	-
numpy.nancumsum	-
numpy.nanmax	cupy.nanmax
numpy.nanmean	cupy.nanmean
numpy.nanmedian	-
numpy.nanmin	cupy.nanmin
numpy.nanpercentile	-
numpy.nanprod	cupy.nanprod
numpy.nanquantile	-
numpy.nanstd	cupy.nanstd
numpy.nansum	cupy.nansum
numpy.nanvar	cupy.nanvar
numpy.ndfromtxt	-
numpy.ndim	-
numpy.negative	cupy.negative
numpy.nested_iters	-
numpy.nextafter	cupy.nextafter
numpy.nonzero	cupy.nonzero
numpy.not_equal	cupy.not_equal
numpy.nper	-
numpy.npv	-
numpy.obj2sctype	cupy.obj2sctype (alias of numpy.obj2sctype)
numpy.ones	cupy.ones
numpy.ones_like	cupy.ones_like
numpy.outer	cupy.outer
numpy.packbits	cupy.packbits
numpy.pad	cupy.pad
numpy.partition	cupy.partition
numpy.percentile	cupy.percentile
numpy.piecewise	-
numpy.place	cupy.place
numpy.pmt	-
numpy.poly	-
numpy.polyadd	-
numpy.polyder	-
numpy.polydiv	-
numpy.polyfit	-
numpy.polyint	-
numpy.polymul	-
numpy.polysub	-
numpy.polyval	-
numpy.positive	-
numpy.power	cupy.power
numpy.ppmt	-
numpy.printoptions	-
numpy.prod	cupy.prod
numpy.product	-
numpy.promote_types	cupy.promote_types (alias of numpy.promote_types)
numpy.ptp	-
numpy.put	cupy.put
numpy.put_along_axis	-
numpy.putmask	-
numpy.pv	-
numpy.quantile	-
numpy.rad2deg	cupy.rad2deg
numpy.radians	cupy.radians
numpy.rate	-
numpy.ravel	cupy.ravel
numpy.ravel_multi_index	-
numpy.real	cupy.real
numpy.real_if_close	-
numpy.recfromcsv	-
numpy.recfromtxt	-
numpy.reciprocal	cupy.reciprocal
numpy.remainder	cupy.remainder
numpy.repeat	cupy.repeat
numpy.require	-
numpy.reshape	cupy.reshape
numpy.resize	-
numpy.result_type	cupy.result_type
numpy.right_shift	cupy.right_shift
numpy.rint	cupy.rint
numpy.roll	cupy.roll
numpy.rollaxis	cupy.rollaxis
numpy.roots	-
numpy.rot90	cupy.rot90
numpy.round	-
numpy.round_	cupy.round_
numpy.row_stack	-
numpy.safe_eval	-
numpy.save	cupy.save
numpy.savetxt	-
numpy.savez	cupy.savez
numpy.savez_compressed	cupy.savez_compressed
numpy.sctype2char	cupy.sctype2char (alias of numpy.sctype2char)
numpy.searchsorted	cupy.searchsorted
numpy.select	-
numpy.set_numeric_ops	-
numpy.set_printoptions	-
numpy.set_string_function	-
numpy.setbufsize	-
numpy.setdiff1d	-
numpy.seterr	-
numpy.seterrcall	-
numpy.seterrobj	-
numpy.setxor1d	-
numpy.shape	-
numpy.shares_memory	cupy.shares_memory
numpy.show_config	cupy.show_config
numpy.sign	cupy.sign
numpy.signbit	cupy.signbit
numpy.sin	cupy.sin
numpy.sinc	cupy.sinc
numpy.sinh	cupy.sinh
numpy.size	cupy.size
numpy.sometrue	-
numpy.sort	cupy.sort
numpy.sort_complex	-
numpy.source	-
numpy.spacing	-
numpy.split	cupy.split
numpy.sqrt	cupy.sqrt
numpy.square	cupy.square
numpy.squeeze	cupy.squeeze
numpy.stack	cupy.stack
numpy.std	cupy.std
numpy.subtract	cupy.subtract
numpy.sum	cupy.sum
numpy.swapaxes	cupy.swapaxes
numpy.take	cupy.take
numpy.take_along_axis	cupy.take_along_axis
numpy.tan	cupy.tan
numpy.tanh	cupy.tanh
numpy.tensordot	cupy.tensordot
numpy.tile	cupy.tile
numpy.trace	cupy.trace
numpy.transpose	cupy.transpose
numpy.trapz	-
numpy.tri	cupy.tri
numpy.tril	cupy.tril
numpy.tril_indices	-
numpy.tril_indices_from	-
numpy.trim_zeros	-
numpy.triu	cupy.triu
numpy.triu_indices	-
numpy.triu_indices_from	-
numpy.true_divide	cupy.true_divide
numpy.trunc	cupy.trunc
numpy.typename	cupy.typename (alias of numpy.typename)
numpy.union1d	-
numpy.unique	cupy.unique
numpy.unpackbits	cupy.unpackbits
numpy.unravel_index	cupy.unravel_index
numpy.unwrap	cupy.unwrap
numpy.vander	-
numpy.var	cupy.var
numpy.vdot	cupy.vdot
numpy.vsplit	cupy.vsplit
numpy.vstack	cupy.vstack
numpy.where	cupy.where
numpy.who	-
numpy.zeros	cupy.zeros
numpy.zeros_like	cupy.zeros_like

Multi-Dimensional Array

NumPy	CuPy
numpy.ndarray.all()	cupy.ndarray.all()
numpy.ndarray.any()	cupy.ndarray.any()
numpy.ndarray.argmax()	cupy.ndarray.argmax()
numpy.ndarray.argmin()	cupy.ndarray.argmin()
numpy.ndarray.argpartition()	cupy.ndarray.argpartition()
numpy.ndarray.argsort()	cupy.ndarray.argsort()
numpy.ndarray.astype()	cupy.ndarray.astype()
numpy.ndarray.byteswap()	-
numpy.ndarray.choose()	cupy.ndarray.choose()
numpy.ndarray.clip()	cupy.ndarray.clip()
numpy.ndarray.compress()	-
numpy.ndarray.conj()	cupy.ndarray.conj()
numpy.ndarray.conjugate()	-
numpy.ndarray.copy()	cupy.ndarray.copy()
numpy.ndarray.cumprod()	cupy.ndarray.cumprod()
numpy.ndarray.cumsum()	cupy.ndarray.cumsum()
numpy.ndarray.diagonal()	cupy.ndarray.diagonal()
numpy.ndarray.dot()	cupy.ndarray.dot()
numpy.ndarray.dump()	cupy.ndarray.dump()
numpy.ndarray.dumps()	cupy.ndarray.dumps()
numpy.ndarray.fill()	cupy.ndarray.fill()
numpy.ndarray.flatten()	cupy.ndarray.flatten()
numpy.ndarray.getfield()	-
numpy.ndarray.item()	cupy.ndarray.item()
numpy.ndarray.itemset()	-
numpy.ndarray.max()	cupy.ndarray.max()
numpy.ndarray.mean()	cupy.ndarray.mean()
numpy.ndarray.min()	cupy.ndarray.min()
numpy.ndarray.newbyteorder()	-
numpy.ndarray.nonzero()	cupy.ndarray.nonzero()
numpy.ndarray.partition()	cupy.ndarray.partition()
numpy.ndarray.prod()	cupy.ndarray.prod()
numpy.ndarray.ptp()	-
numpy.ndarray.put()	cupy.ndarray.put()
numpy.ndarray.ravel()	cupy.ndarray.ravel()
numpy.ndarray.repeat()	cupy.ndarray.repeat()
numpy.ndarray.reshape()	cupy.ndarray.reshape()
numpy.ndarray.resize()	-
numpy.ndarray.round()	cupy.ndarray.round()
numpy.ndarray.searchsorted()	-
numpy.ndarray.setfield()	-
numpy.ndarray.setflags()	-
numpy.ndarray.sort()	cupy.ndarray.sort()
numpy.ndarray.squeeze()	cupy.ndarray.squeeze()
numpy.ndarray.std()	cupy.ndarray.std()
numpy.ndarray.sum()	cupy.ndarray.sum()
numpy.ndarray.swapaxes()	cupy.ndarray.swapaxes()
numpy.ndarray.take()	cupy.ndarray.take()
numpy.ndarray.tobytes()	cupy.ndarray.tobytes()
numpy.ndarray.tofile()	cupy.ndarray.tofile()
numpy.ndarray.tolist()	cupy.ndarray.tolist()
numpy.ndarray.tostring()	-
numpy.ndarray.trace()	cupy.ndarray.trace()
numpy.ndarray.transpose()	cupy.ndarray.transpose()
numpy.ndarray.var()	cupy.ndarray.var()
numpy.ndarray.view()	cupy.ndarray.view()

Linear Algebra

NumPy	CuPy
numpy.linalg.cholesky	cupy.linalg.cholesky
numpy.linalg.cond	-
numpy.linalg.det	cupy.linalg.det
numpy.linalg.eig	-
numpy.linalg.eigh	cupy.linalg.eigh
numpy.linalg.eigvals	-
numpy.linalg.eigvalsh	cupy.linalg.eigvalsh
numpy.linalg.inv	cupy.linalg.inv
numpy.linalg.lstsq	cupy.linalg.lstsq
numpy.linalg.matrix_power	cupy.linalg.matrix_power
numpy.linalg.matrix_rank	cupy.linalg.matrix_rank
numpy.linalg.multi_dot	-
numpy.linalg.norm	cupy.linalg.norm
numpy.linalg.pinv	cupy.linalg.pinv
numpy.linalg.qr	cupy.linalg.qr
numpy.linalg.slogdet	cupy.linalg.slogdet
numpy.linalg.solve	cupy.linalg.solve
numpy.linalg.svd	cupy.linalg.svd
numpy.linalg.tensorinv	cupy.linalg.tensorinv
numpy.linalg.tensorsolve	cupy.linalg.tensorsolve

Discrete Fourier Transform

NumPy	CuPy
numpy.fft.fft	cupy.fft.fft
numpy.fft.fft2	cupy.fft.fft2
numpy.fft.fftfreq	cupy.fft.fftfreq
numpy.fft.fftn	cupy.fft.fftn
numpy.fft.fftshift	cupy.fft.fftshift
numpy.fft.hfft	cupy.fft.hfft
numpy.fft.ifft	cupy.fft.ifft
numpy.fft.ifft2	cupy.fft.ifft2
numpy.fft.ifftn	cupy.fft.ifftn
numpy.fft.ifftshift	cupy.fft.ifftshift
numpy.fft.ihfft	cupy.fft.ihfft
numpy.fft.irfft	cupy.fft.irfft
numpy.fft.irfft2	cupy.fft.irfft2
numpy.fft.irfftn	cupy.fft.irfftn
numpy.fft.rfft	cupy.fft.rfft
numpy.fft.rfft2	cupy.fft.rfft2
numpy.fft.rfftfreq	cupy.fft.rfftfreq
numpy.fft.rfftn	cupy.fft.rfftn

Random Sampling

NumPy	CuPy
numpy.random.beta	cupy.random.beta
numpy.random.binomial	cupy.random.binomial
numpy.random.bytes	cupy.random.bytes
numpy.random.chisquare	cupy.random.chisquare
numpy.random.choice	cupy.random.choice
numpy.random.default_rng	-
numpy.random.dirichlet	cupy.random.dirichlet
numpy.random.exponential	cupy.random.exponential
numpy.random.f	cupy.random.f
numpy.random.gamma	cupy.random.gamma
numpy.random.geometric	cupy.random.geometric
numpy.random.get_state	-
numpy.random.gumbel	cupy.random.gumbel
numpy.random.hypergeometric	cupy.random.hypergeometric
numpy.random.laplace	cupy.random.laplace
numpy.random.logistic	cupy.random.logistic
numpy.random.lognormal	cupy.random.lognormal
numpy.random.logseries	cupy.random.logseries
numpy.random.multinomial	cupy.random.multinomial
numpy.random.multivariate_normal	cupy.random.multivariate_normal
numpy.random.negative_binomial	cupy.random.negative_binomial
numpy.random.noncentral_chisquare	cupy.random.noncentral_chisquare
numpy.random.noncentral_f	cupy.random.noncentral_f
numpy.random.normal	cupy.random.normal
numpy.random.pareto	cupy.random.pareto
numpy.random.permutation	cupy.random.permutation
numpy.random.poisson	cupy.random.poisson
numpy.random.power	cupy.random.power
numpy.random.rand	cupy.random.rand
numpy.random.randint	cupy.random.randint
numpy.random.randn	cupy.random.randn
numpy.random.random	cupy.random.random
numpy.random.random_integers	cupy.random.random_integers
numpy.random.random_sample	cupy.random.random_sample
numpy.random.ranf	cupy.random.ranf
numpy.random.rayleigh	cupy.random.rayleigh
numpy.random.sample	cupy.random.sample
numpy.random.seed	cupy.random.seed
numpy.random.set_state	-
numpy.random.shuffle	cupy.random.shuffle
numpy.random.standard_cauchy	cupy.random.standard_cauchy
numpy.random.standard_exponential	cupy.random.standard_exponential
numpy.random.standard_gamma	cupy.random.standard_gamma
numpy.random.standard_normal	cupy.random.standard_normal
numpy.random.standard_t	cupy.random.standard_t
numpy.random.triangular	cupy.random.triangular
numpy.random.uniform	cupy.random.uniform
numpy.random.vonmises	cupy.random.vonmises
numpy.random.wald	cupy.random.wald
numpy.random.weibull	cupy.random.weibull
numpy.random.zipf	cupy.random.zipf

SciPy / CuPy APIs

Discrete Fourier Transform

SciPy	CuPy
scipy.fftpack.cc_diff	-
scipy.fftpack.cs_diff	-
scipy.fftpack.dct	-
scipy.fftpack.dctn	-
scipy.fftpack.diff	-
scipy.fftpack.dst	-
scipy.fftpack.dstn	-
scipy.fftpack.fft	cupyx.scipy.fftpack.fft
scipy.fftpack.fft2	cupyx.scipy.fftpack.fft2
scipy.fftpack.fftfreq	-
scipy.fftpack.fftn	cupyx.scipy.fftpack.fftn
scipy.fftpack.fftshift	-
scipy.fftpack.hilbert	-
scipy.fftpack.idct	-
scipy.fftpack.idctn	-
scipy.fftpack.idst	-
scipy.fftpack.idstn	-
scipy.fftpack.ifft	cupyx.scipy.fftpack.ifft
scipy.fftpack.ifft2	cupyx.scipy.fftpack.ifft2
scipy.fftpack.ifftn	cupyx.scipy.fftpack.ifftn
scipy.fftpack.ifftshift	-
scipy.fftpack.ihilbert	-
scipy.fftpack.irfft	cupyx.scipy.fftpack.irfft
scipy.fftpack.itilbert	-
scipy.fftpack.next_fast_len	-
scipy.fftpack.rfft	cupyx.scipy.fftpack.rfft
scipy.fftpack.rfftfreq	-
scipy.fftpack.sc_diff	-
scipy.fftpack.shift	-
scipy.fftpack.ss_diff	-
scipy.fftpack.tilbert	-

Sparse Matrices

SciPy	CuPy
scipy.sparse.block_diag	-
scipy.sparse.bmat	-
scipy.sparse.diags	cupyx.scipy.sparse.diags
scipy.sparse.eye	cupyx.scipy.sparse.eye
scipy.sparse.find	-
scipy.sparse.hstack	-
scipy.sparse.identity	cupyx.scipy.sparse.identity
scipy.sparse.issparse	cupyx.scipy.sparse.issparse
scipy.sparse.isspmatrix	cupyx.scipy.sparse.isspmatrix
scipy.sparse.isspmatrix_bsr	-
scipy.sparse.isspmatrix_coo	cupyx.scipy.sparse.isspmatrix_coo
scipy.sparse.isspmatrix_csc	cupyx.scipy.sparse.isspmatrix_csc
scipy.sparse.isspmatrix_csr	cupyx.scipy.sparse.isspmatrix_csr
scipy.sparse.isspmatrix_dia	cupyx.scipy.sparse.isspmatrix_dia
scipy.sparse.isspmatrix_dok	-
scipy.sparse.isspmatrix_lil	-
scipy.sparse.kron	-
scipy.sparse.kronsum	-
scipy.sparse.load_npz	-
scipy.sparse.rand	cupyx.scipy.sparse.rand
scipy.sparse.random	cupyx.scipy.sparse.random
scipy.sparse.save_npz	-
scipy.sparse.spdiags	cupyx.scipy.sparse.spdiags
scipy.sparse.tril	-
scipy.sparse.triu	-
scipy.sparse.vstack	-

Sparse Linear Algebra

SciPy	CuPy
scipy.sparse.linalg.aslinearoperator	-
scipy.sparse.linalg.bicg	-
scipy.sparse.linalg.bicgstab	-
scipy.sparse.linalg.cg	-
scipy.sparse.linalg.cgs	-
scipy.sparse.linalg.eigs	-
scipy.sparse.linalg.eigsh	-
scipy.sparse.linalg.expm	-
scipy.sparse.linalg.expm_multiply	-
scipy.sparse.linalg.factorized	-
scipy.sparse.linalg.gcrotmk	-
scipy.sparse.linalg.gmres	-
scipy.sparse.linalg.inv	-
scipy.sparse.linalg.lgmres	-
scipy.sparse.linalg.lobpcg	-
scipy.sparse.linalg.lsmr	-
scipy.sparse.linalg.lsqr	cupyx.scipy.sparse.linalg.lsqr
scipy.sparse.linalg.minres	-
scipy.sparse.linalg.norm	-
scipy.sparse.linalg.onenormest	-
scipy.sparse.linalg.qmr	-
scipy.sparse.linalg.spilu	-
scipy.sparse.linalg.splu	-
scipy.sparse.linalg.spsolve	-
scipy.sparse.linalg.spsolve_triangular	-
scipy.sparse.linalg.svds	-
scipy.sparse.linalg.use_solver	-

Advanced Linear Algebra

SciPy	CuPy
scipy.linalg.block_diag	-
scipy.linalg.cdf2rdf	-
scipy.linalg.cho_factor	-
scipy.linalg.cho_solve	-
scipy.linalg.cho_solve_banded	-
scipy.linalg.cholesky_banded	-
scipy.linalg.circulant	-
scipy.linalg.clarkson_woodruff_transform	-
scipy.linalg.companion	-
scipy.linalg.coshm	-
scipy.linalg.cosm	-
scipy.linalg.dft	-
scipy.linalg.diagsvd	-
scipy.linalg.eig_banded	-
scipy.linalg.eigh_tridiagonal	-
scipy.linalg.eigvals_banded	-
scipy.linalg.eigvalsh_tridiagonal	-
scipy.linalg.expm	-
scipy.linalg.expm_cond	-
scipy.linalg.expm_frechet	-
scipy.linalg.fiedler	-
scipy.linalg.fiedler_companion	-
scipy.linalg.find_best_blas_type	-
scipy.linalg.fractional_matrix_power	-
scipy.linalg.funm	-
scipy.linalg.get_blas_funcs	-
scipy.linalg.get_lapack_funcs	-
scipy.linalg.hadamard	-
scipy.linalg.hankel	-
scipy.linalg.helmert	-
scipy.linalg.hessenberg	-
scipy.linalg.hilbert	-
scipy.linalg.invhilbert	-
scipy.linalg.invpascal	-
scipy.linalg.kron	-
scipy.linalg.ldl	-
scipy.linalg.leslie	-
scipy.linalg.logm	-
scipy.linalg.lu	-
scipy.linalg.lu_factor	cupyx.scipy.linalg.lu_factor
scipy.linalg.lu_solve	cupyx.scipy.linalg.lu_solve
scipy.linalg.matrix_balance	-
scipy.linalg.null_space	-
scipy.linalg.ordqz	-
scipy.linalg.orth	-
scipy.linalg.orthogonal_procrustes	-
scipy.linalg.pascal	-
scipy.linalg.pinv2	-
scipy.linalg.pinvh	-
scipy.linalg.polar	-
scipy.linalg.qr_delete	-
scipy.linalg.qr_insert	-
scipy.linalg.qr_multiply	-
scipy.linalg.qr_update	-
scipy.linalg.qz	-
scipy.linalg.rq	-
scipy.linalg.rsf2csf	-
scipy.linalg.schur	-
scipy.linalg.signm	-
scipy.linalg.sinhm	-
scipy.linalg.sinm	-
scipy.linalg.solve_banded	-
scipy.linalg.solve_circulant	-
scipy.linalg.solve_continuous_are	-
scipy.linalg.solve_continuous_lyapunov	-
scipy.linalg.solve_discrete_are	-
scipy.linalg.solve_discrete_lyapunov	-
scipy.linalg.solve_lyapunov	-
scipy.linalg.solve_sylvester	-
scipy.linalg.solve_toeplitz	-
scipy.linalg.solve_triangular	cupyx.scipy.linalg.solve_triangular
scipy.linalg.solveh_banded	-
scipy.linalg.sqrtm	-
scipy.linalg.subspace_angles	-
scipy.linalg.svdvals	-
scipy.linalg.tanhm	-
scipy.linalg.tanm	-
scipy.linalg.toeplitz	-
scipy.linalg.tri	-
scipy.linalg.tril	-
scipy.linalg.triu	-

Multidimensional Image Processing

SciPy	CuPy
scipy.ndimage.affine_transform	cupyx.scipy.ndimage.affine_transform
scipy.ndimage.binary_closing	-
scipy.ndimage.binary_dilation	-
scipy.ndimage.binary_erosion	-
scipy.ndimage.binary_fill_holes	-
scipy.ndimage.binary_hit_or_miss	-
scipy.ndimage.binary_opening	-
scipy.ndimage.binary_propagation	-
scipy.ndimage.black_tophat	-
scipy.ndimage.center_of_mass	-
scipy.ndimage.convolve	cupyx.scipy.ndimage.convolve
scipy.ndimage.convolve1d	-
scipy.ndimage.correlate	cupyx.scipy.ndimage.correlate
scipy.ndimage.correlate1d	-
scipy.ndimage.distance_transform_bf	-
scipy.ndimage.distance_transform_cdt	-
scipy.ndimage.distance_transform_edt	-
scipy.ndimage.extrema	-
scipy.ndimage.find_objects	-
scipy.ndimage.fourier_ellipsoid	-
scipy.ndimage.fourier_gaussian	-
scipy.ndimage.fourier_shift	-
scipy.ndimage.fourier_uniform	-
scipy.ndimage.gaussian_filter	-
scipy.ndimage.gaussian_filter1d	-
scipy.ndimage.gaussian_gradient_magnitude	-
scipy.ndimage.gaussian_laplace	-
scipy.ndimage.generate_binary_structure	-
scipy.ndimage.generic_filter	-
scipy.ndimage.generic_filter1d	-
scipy.ndimage.generic_gradient_magnitude	-
scipy.ndimage.generic_laplace	-
scipy.ndimage.geometric_transform	-
scipy.ndimage.grey_closing	-
scipy.ndimage.grey_dilation	-
scipy.ndimage.grey_erosion	-
scipy.ndimage.grey_opening	-
scipy.ndimage.histogram	-
scipy.ndimage.iterate_structure	-
scipy.ndimage.label	-
scipy.ndimage.labeled_comprehension	-
scipy.ndimage.laplace	-
scipy.ndimage.map_coordinates	cupyx.scipy.ndimage.map_coordinates
scipy.ndimage.maximum	-
scipy.ndimage.maximum_filter	-
scipy.ndimage.maximum_filter1d	-
scipy.ndimage.maximum_position	-
scipy.ndimage.mean	-
scipy.ndimage.median	-
scipy.ndimage.median_filter	-
scipy.ndimage.minimum	-
scipy.ndimage.minimum_filter	-
scipy.ndimage.minimum_filter1d	-
scipy.ndimage.minimum_position	-
scipy.ndimage.morphological_gradient	-
scipy.ndimage.morphological_laplace	-
scipy.ndimage.percentile_filter	-
scipy.ndimage.prewitt	-
scipy.ndimage.rank_filter	-
scipy.ndimage.rotate	cupyx.scipy.ndimage.rotate
scipy.ndimage.shift	cupyx.scipy.ndimage.shift
scipy.ndimage.sobel	-
scipy.ndimage.spline_filter	-
scipy.ndimage.spline_filter1d	-
scipy.ndimage.standard_deviation	-
scipy.ndimage.sum	-
scipy.ndimage.uniform_filter	-
scipy.ndimage.uniform_filter1d	-
scipy.ndimage.variance	-
scipy.ndimage.watershed_ift	-
scipy.ndimage.white_tophat	-
scipy.ndimage.zoom	cupyx.scipy.ndimage.zoom

Special Functions

SciPy	CuPy
scipy.special.agm	-
scipy.special.ai_zeros	-
scipy.special.airy	-
scipy.special.airye	-
scipy.special.assoc_laguerre	-
scipy.special.bdtr	-
scipy.special.bdtrc	-
scipy.special.bdtri	-
scipy.special.bdtrik	-
scipy.special.bdtrin	-
scipy.special.bei	-
scipy.special.bei_zeros	-
scipy.special.beip	-
scipy.special.beip_zeros	-
scipy.special.ber	-
scipy.special.ber_zeros	-
scipy.special.bernoulli	-
scipy.special.berp	-
scipy.special.berp_zeros	-
scipy.special.besselpoly	-
scipy.special.beta	-
scipy.special.betainc	-
scipy.special.betaincinv	-
scipy.special.betaln	-
scipy.special.bi_zeros	-
scipy.special.binom	-
scipy.special.boxcox	-
scipy.special.boxcox1p	-
scipy.special.btdtr	-
scipy.special.btdtri	-
scipy.special.btdtria	-
scipy.special.btdtrib	-
scipy.special.c_roots	-
scipy.special.cbrt	-
scipy.special.cg_roots	-
scipy.special.chdtr	-
scipy.special.chdtrc	-
scipy.special.chdtri	-
scipy.special.chdtriv	-
scipy.special.chebyc	-
scipy.special.chebys	-
scipy.special.chebyt	-
scipy.special.chebyu	-
scipy.special.chndtr	-
scipy.special.chndtridf	-
scipy.special.chndtrinc	-
scipy.special.chndtrix	-
scipy.special.clpmn	-
scipy.special.comb	-
scipy.special.cosdg	-
scipy.special.cosm1	-
scipy.special.cotdg	-
scipy.special.dawsn	-
scipy.special.digamma	cupyx.scipy.special.digamma
scipy.special.diric	-
scipy.special.ellip_harm	-
scipy.special.ellip_harm_2	-
scipy.special.ellip_normal	-
scipy.special.ellipe	-
scipy.special.ellipeinc	-
scipy.special.ellipj	-
scipy.special.ellipk	-
scipy.special.ellipkinc	-
scipy.special.ellipkm1	-
scipy.special.entr	-
scipy.special.erf	cupyx.scipy.special.erf
scipy.special.erf_zeros	-
scipy.special.erfc	cupyx.scipy.special.erfc
scipy.special.erfcinv	cupyx.scipy.special.erfcinv
scipy.special.erfcx	cupyx.scipy.special.erfcx
scipy.special.erfi	-
scipy.special.erfinv	cupyx.scipy.special.erfinv
scipy.special.euler	-
scipy.special.eval_chebyc	-
scipy.special.eval_chebys	-
scipy.special.eval_chebyt	-
scipy.special.eval_chebyu	-
scipy.special.eval_gegenbauer	-
scipy.special.eval_genlaguerre	-
scipy.special.eval_hermite	-
scipy.special.eval_hermitenorm	-
scipy.special.eval_jacobi	-
scipy.special.eval_laguerre	-
scipy.special.eval_legendre	-
scipy.special.eval_sh_chebyt	-
scipy.special.eval_sh_chebyu	-
scipy.special.eval_sh_jacobi	-
scipy.special.eval_sh_legendre	-
scipy.special.exp1	-
scipy.special.exp10	-
scipy.special.exp2	-
scipy.special.expi	-
scipy.special.expit	-
scipy.special.expm1	-
scipy.special.expn	-
scipy.special.exprel	-
scipy.special.factorial	-
scipy.special.factorial2	-
scipy.special.factorialk	-
scipy.special.fdtr	-
scipy.special.fdtrc	-
scipy.special.fdtri	-
scipy.special.fdtridfd	-
scipy.special.fresnel	-
scipy.special.fresnel_zeros	-
scipy.special.fresnelc_zeros	-
scipy.special.fresnels_zeros	-
scipy.special.gamma	cupyx.scipy.special.gamma
scipy.special.gammainc	-
scipy.special.gammaincc	-
scipy.special.gammainccinv	-
scipy.special.gammaincinv	-
scipy.special.gammaln	cupyx.scipy.special.gammaln
scipy.special.gammasgn	-
scipy.special.gdtr	-
scipy.special.gdtrc	-
scipy.special.gdtria	-
scipy.special.gdtrib	-
scipy.special.gdtrix	-
scipy.special.gegenbauer	-
scipy.special.genlaguerre	-
scipy.special.geterr	-
scipy.special.h1vp	-
scipy.special.h2vp	-
scipy.special.h_roots	-
scipy.special.hankel1	-
scipy.special.hankel1e	-
scipy.special.hankel2	-
scipy.special.hankel2e	-
scipy.special.he_roots	-
scipy.special.hermite	-
scipy.special.hermitenorm	-
scipy.special.huber	-
scipy.special.hyp0f1	-
scipy.special.hyp1f1	-
scipy.special.hyp2f1	-
scipy.special.hyperu	-
scipy.special.i0	cupyx.scipy.special.i0
scipy.special.i0e	-
scipy.special.i1	cupyx.scipy.special.i1
scipy.special.i1e	-
scipy.special.inv_boxcox	-
scipy.special.inv_boxcox1p	-
scipy.special.it2i0k0	-
scipy.special.it2j0y0	-
scipy.special.it2struve0	-
scipy.special.itairy	-
scipy.special.iti0k0	-
scipy.special.itj0y0	-
scipy.special.itmodstruve0	-
scipy.special.itstruve0	-
scipy.special.iv	-
scipy.special.ive	-
scipy.special.ivp	-
scipy.special.j0	cupyx.scipy.special.j0
scipy.special.j1	cupyx.scipy.special.j1
scipy.special.j_roots	-
scipy.special.jacobi	-
scipy.special.jn	-
scipy.special.jn_zeros	-
scipy.special.jnjnp_zeros	-
scipy.special.jnp_zeros	-
scipy.special.jnyn_zeros	-
scipy.special.js_roots	-
scipy.special.jv	-
scipy.special.jve	-
scipy.special.jvp	-
scipy.special.k0	-
scipy.special.k0e	-
scipy.special.k1	-
scipy.special.k1e	-
scipy.special.kei	-
scipy.special.kei_zeros	-
scipy.special.keip	-
scipy.special.keip_zeros	-
scipy.special.kelvin	-
scipy.special.kelvin_zeros	-
scipy.special.ker	-
scipy.special.ker_zeros	-
scipy.special.kerp	-
scipy.special.kerp_zeros	-
scipy.special.kl_div	-
scipy.special.kn	-
scipy.special.kolmogi	-
scipy.special.kolmogorov	-
scipy.special.kv	-
scipy.special.kve	-
scipy.special.kvp	-
scipy.special.l_roots	-
scipy.special.la_roots	-
scipy.special.laguerre	-
scipy.special.lambertw	-
scipy.special.legendre	-
scipy.special.lmbda	-
scipy.special.log1p	-
scipy.special.log_ndtr	-
scipy.special.loggamma	-
scipy.special.logit	-
scipy.special.logsumexp	-
scipy.special.lpmn	-
scipy.special.lpmv	-
scipy.special.lpn	-
scipy.special.lqmn	-
scipy.special.lqn	-
scipy.special.mathieu_a	-
scipy.special.mathieu_b	-
scipy.special.mathieu_cem	-
scipy.special.mathieu_even_coef	-
scipy.special.mathieu_modcem1	-
scipy.special.mathieu_modcem2	-
scipy.special.mathieu_modsem1	-
scipy.special.mathieu_modsem2	-
scipy.special.mathieu_odd_coef	-
scipy.special.mathieu_sem	-
scipy.special.modfresnelm	-
scipy.special.modfresnelp	-
scipy.special.modstruve	-
scipy.special.multigammaln	-
scipy.special.nbdtr	-
scipy.special.nbdtrc	-
scipy.special.nbdtri	-
scipy.special.nbdtrik	-
scipy.special.nbdtrin	-
scipy.special.ncfdtr	-
scipy.special.ncfdtri	-
scipy.special.ncfdtridfd	-
scipy.special.ncfdtridfn	-
scipy.special.ncfdtrinc	-
scipy.special.nctdtr	-
scipy.special.nctdtridf	-
scipy.special.nctdtrinc	-
scipy.special.nctdtrit	-
scipy.special.ndtr	cupyx.scipy.special.ndtr
scipy.special.ndtri	-
scipy.special.nrdtrimn	-
scipy.special.nrdtrisd	-
scipy.special.obl_ang1	-
scipy.special.obl_ang1_cv	-
scipy.special.obl_cv	-
scipy.special.obl_cv_seq	-
scipy.special.obl_rad1	-
scipy.special.obl_rad1_cv	-
scipy.special.obl_rad2	-
scipy.special.obl_rad2_cv	-
scipy.special.owens_t	-
scipy.special.p_roots	-
scipy.special.pbdn_seq	-
scipy.special.pbdv	-
scipy.special.pbdv_seq	-
scipy.special.pbvv	-
scipy.special.pbvv_seq	-
scipy.special.pbwa	-
scipy.special.pdtr	-
scipy.special.pdtrc	-
scipy.special.pdtri	-
scipy.special.pdtrik	-
scipy.special.perm	-
scipy.special.poch	-
scipy.special.polygamma	cupyx.scipy.special.polygamma
scipy.special.pro_ang1	-
scipy.special.pro_ang1_cv	-
scipy.special.pro_cv	-
scipy.special.pro_cv_seq	-
scipy.special.pro_rad1	-
scipy.special.pro_rad1_cv	-
scipy.special.pro_rad2	-
scipy.special.pro_rad2_cv	-
scipy.special.ps_roots	-
scipy.special.pseudo_huber	-
scipy.special.psi	-
scipy.special.radian	-
scipy.special.rel_entr	-
scipy.special.rgamma	-
scipy.special.riccati_jn	-
scipy.special.riccati_yn	-
scipy.special.roots_chebyc	-
scipy.special.roots_chebys	-
scipy.special.roots_chebyt	-
scipy.special.roots_chebyu	-
scipy.special.roots_gegenbauer	-
scipy.special.roots_genlaguerre	-
scipy.special.roots_hermite	-
scipy.special.roots_hermitenorm	-
scipy.special.roots_jacobi	-
scipy.special.roots_laguerre	-
scipy.special.roots_legendre	-
scipy.special.roots_sh_chebyt	-
scipy.special.roots_sh_chebyu	-
scipy.special.roots_sh_jacobi	-
scipy.special.roots_sh_legendre	-
scipy.special.round	-
scipy.special.s_roots	-
scipy.special.seterr	-
scipy.special.sh_chebyt	-
scipy.special.sh_chebyu	-
scipy.special.sh_jacobi	-
scipy.special.sh_legendre	-
scipy.special.shichi	-
scipy.special.sici	-
scipy.special.sinc	-
scipy.special.sindg	-
scipy.special.smirnov	-
scipy.special.smirnovi	-
scipy.special.softmax	-
scipy.special.spence	-
scipy.special.sph_harm	-
scipy.special.spherical_in	-
scipy.special.spherical_jn	-
scipy.special.spherical_kn	-
scipy.special.spherical_yn	-
scipy.special.stdtr	-
scipy.special.stdtridf	-
scipy.special.stdtrit	-
scipy.special.struve	-
scipy.special.t_roots	-
scipy.special.tandg	-
scipy.special.tklmbda	-
scipy.special.ts_roots	-
scipy.special.u_roots	-
scipy.special.us_roots	-
scipy.special.voigt_profile	-
scipy.special.wofz	-
scipy.special.wrightomega	-
scipy.special.xlog1py	-
scipy.special.xlogy	-
scipy.special.y0	cupyx.scipy.special.y0
scipy.special.y0_zeros	-
scipy.special.y1	cupyx.scipy.special.y1
scipy.special.y1_zeros	-
scipy.special.y1p_zeros	-
scipy.special.yn	-
scipy.special.yn_zeros	-
scipy.special.ynp_zeros	-
scipy.special.yv	-
scipy.special.yve	-
scipy.special.yvp	-
scipy.special.zeta	cupyx.scipy.special.zeta
scipy.special.zetac	-

API Compatibility Policy

This document expresses the design policy on compatibilities of CuPy APIs. Development team should obey this policy on deciding to add, extend, and change APIs and their behaviors.

This document is written for both users and developers. Users can decide the level of dependencies on CuPy’s implementations in their codes based on this document. Developers should read through this document before creating pull requests that contain changes on the interface. Note that this document may contain ambiguities on the level of supported compatibilities.

Versioning and Backward Compatibilities

The updates of CuPy are classified into three levels: major, minor, and revision. These types have distinct levels of backward compatibilities.

• Major update contains disruptive changes that break the backward compatibility.
• Minor update contains addition and extension to the APIs keeping the supported backward compatibility.
• Revision update contains improvements on the API implementations without changing any API specifications.

Note that we do not support full backward compatibility, which is almost infeasible for Python-based APIs, since there is no way to completely hide the implementation details.

Processes to Break Backward Compatibilities

Deprecation, Dropping, and Its Preparation

Any APIs may be deprecated at some minor updates. In such a case, the deprecation note is added to the API documentation, and the API implementation is changed to fire deprecation warning (if possible). There should be another way to reimplement the same things previously written with the deprecated APIs.

Any APIs may be marked as to be dropped in the future. In such a case, the dropping is stated in the documentation with the major version number on which the API is planned to be dropped, and the API implementation is changed to fire the future warning (if possible).

The actual dropping should be done through the following steps:

• Make the API deprecated. At this point, users should not need the deprecated API in their new application codes.
• After that, mark the API as to be dropped in the future. It must be done in the minor update different from that of the deprecation.
• At the major version announced in the above update, drop the API.

Consequently, it takes at least two minor versions to drop any APIs after the first deprecation.

API Changes and Its Preparation

Any APIs may be marked as to be changed in the future for changes without backward compatibility. In such a case, the change is stated in the documentation with the version number on which the API is planned to be changed, and the API implementation is changed to fire the future warning on the certain usages.

The actual change should be done in the following steps:

• Announce that the API will be changed in the future. At this point, the actual version of change need not be accurate.
• After the announcement, mark the API as to be changed in the future with version number of planned changes. At this point, users should not use the marked API in their new application codes.
• At the major update announced in the above update, change the API.

Supported Backward Compatibility

This section defines backward compatibilities that minor updates must maintain.

Documented Interface

CuPy has the official API documentation. Many applications can be written based on the documented features. We support backward compatibilities of documented features. In other words, codes only based on the documented features run correctly with minor/revision-updated versions.

Developers are encouraged to use apparent names for objects of implementation details. For example, attributes outside of the documented APIs should have one or more underscores at the prefix of their names.

Undocumented behaviors

Behaviors of CuPy implementation not stated in the documentation are undefined. Undocumented behaviors are not guaranteed to be stable between different minor/revision versions.

Minor update may contain changes to undocumented behaviors. For example, suppose an API X is added at the minor update. In the previous version, attempts to use X cause AttributeError. This behavior is not stated in the documentation, so this is undefined. Thus, adding the API X in minor version is permissible.

Revision update may also contain changes to undefined behaviors. Typical example is a bug fix. Another example is an improvement on implementation, which may change the internal object structures not shown in the documentation. As a consequence, even revision updates do not support compatibility of pickling, unless the full layout of pickled objects is clearly documented.

Documentation Error

Compatibility is basically determined based on the documentation, though it sometimes contains errors. It may make the APIs confusing to assume the documentation always stronger than the implementations. We therefore may fix the documentation errors in any updates that may break the compatibility in regard to the documentation.

Note

Developers MUST NOT fix the documentation and implementation of the same functionality at the same time in revision updates as “bug fix”. Such a change completely breaks the backward compatibility. If you want to fix the bugs in both sides, first fix the documentation to fit it into the implementation, and start the API changing procedure described above.

Object Attributes and Properties

Object attributes and properties are sometimes replaced by each other at minor updates. It does not break the user codes, except the codes depend on how the attributes and properties are implemented.

Functions and Methods

Methods may be replaced by callable attributes keeping the compatibility of parameters and return values in minor updates. It does not break the user codes, except the codes depend on how the methods and callable attributes are implemented.

Exceptions and Warnings

The specifications of raising exceptions are considered as a part of standard backward compatibilities. No exception is raised in the future versions with correct usages that the documentation allows, unless the API changing process is completed.

On the other hand, warnings may be added at any minor updates for any APIs. It means minor updates do not keep backward compatibility of warnings.

Installation Compatibility

The installation process is another concern of compatibilities. We support environmental compatibilities in the following ways.

• Any changes of dependent libraries that force modifications on the existing environments must be done in major updates. Such changes include following cases:
  • dropping supported versions of dependent libraries (e.g. dropping cuDNN v2)
  • adding new mandatory dependencies (e.g. adding h5py to setup_requires)
• Supporting optional packages/libraries may be done in minor updates (e.g. supporting h5py in optional features).

Note

The installation compatibility does not guarantee that all the features of CuPy correctly run on supported environments. It may contain bugs that only occurs in certain environments. Such bugs should be fixed in some updates.

Contribution Guide

This is a guide for all contributions to CuPy. The development of CuPy is running on the official repository at GitHub. Anyone that wants to register an issue or to send a pull request should read through this document.

Classification of Contributions

There are several ways to contribute to CuPy community:

1. Registering an issue
2. Sending a pull request (PR)
3. Sending a question to CuPy User Group
4. Open-sourcing an external example
5. Writing a post about CuPy

This document mainly focuses on 1 and 2, though other contributions are also appreciated.

Development Cycle

This section explains the development process of CuPy. Before contributing to CuPy, it is strongly recommended to understand the development cycle.

Versioning

The versioning of CuPy follows PEP 440 and a part of Semantic versioning. The version number consists of three or four parts: X.Y.Zw where X denotes the major version, Y denotes the minor version, Z denotes the revision number, and the optional w denotes the prelease suffix. While the major, minor, and revision numbers follow the rule of semantic versioning, the pre-release suffix follows PEP 440 so that the version string is much friendly with Python eco-system.

Note that a major update basically does not contain compatibility-breaking changes from the last release candidate (RC). This is not a strict rule, though; if there is a critical API bug that we have to fix for the major version, we may add breaking changes to the major version up.

As for the backward compatibility, see API Compatibility Policy.

Release Cycle

The first one is the track of stable versions, which is a series of revision updates for the latest major version. The second one is the track of development versions, which is a series of pre-releases for the upcoming major version.

Consider that X.0.0 is the latest major version and Y.0.0, Z.0.0 are the succeeding major versions. Then, the timeline of the updates is depicted by the following table.

Date	ver X	ver Y	ver Z
0 weeks	X.0.0rc1	–	–
4 weeks	X.0.0	Y.0.0a1	–
8 weeks	X.1.0*	Y.0.0b1	–
12 weeks	X.2.0*	Y.0.0rc1	–
16 weeks	–	Y.0.0	Z.0.0a1

(* These might be revision releases)

The dates shown in the left-most column are relative to the release of X.0.0rc1. In particular, each revision/minor release is made four weeks after the previous one of the same major version, and the pre-release of the upcoming major version is made at the same time. Whether these releases are revision or minor is determined based on the contents of each update.

Note that there are only three stable releases for the versions X.x.x. During the parallel development of Y.0.0 and Z.0.0a1, the version Y is treated as an almost-stable version and Z is treated as a development version.

If there is a critical bug found in X.x.x after stopping the development of version X, we may release a hot-fix for this version at any time.

We create a milestone for each upcoming release at GitHub. The GitHub milestone is basically used for collecting the issues and PRs resolved in the release.

Git Branches

The master branch is used to develop pre-release versions. It means that alpha, beta, and RC updates are developed at the master branch. This branch contains the most up-to-date source tree that includes features newly added after the latest major version.

The stable version is developed at the individual branch named as vN where “N” reflects the version number (we call it a versioned branch). For example, v1.0.0, v1.0.1, and v1.0.2 will be developed at the v1 branch.

Notes for contributors: When you send a pull request, you basically have to send it to the master branch. If the change can also be applied to the stable version, a core team member will apply the same change to the stable version so that the change is also included in the next revision update.

If the change is only applicable to the stable version and not to the master branch, please send it to the versioned branch. We basically only accept changes to the latest versioned branch (where the stable version is developed) unless the fix is critical.

If you want to make a new feature of the master branch available in the current stable version, please send a backport PR to the stable version (the latest vN branch). See the next section for details.

Note: a change that can be applied to both branches should be sent to the master branch. Each release of the stable version is also merged to the development version so that the change is also reflected to the next major version.

Feature Backport PRs

We basically do not backport any new features of the development version to the stable versions. If you desire to include the feature to the current stable version and you can work on the backport work, we welcome such a contribution. In such a case, you have to send a backport PR to the latest vN branch. Note that we do not accept any feature backport PRs to older versions because we are not running quality assurance workflows (e.g. CI) for older versions so that we cannot ensure that the PR is correctly ported.

There are some rules on sending a backport PR.

• Start the PR title from the prefix [backport].
• Clarify the original PR number in the PR description (something like “This is a backport of #XXXX”).
• (optional) Write to the PR description the motivation of backporting the feature to the stable version.

Please follow these rules when you create a feature backport PR.

Note: PRs that do not include any changes/additions to APIs (e.g. bug fixes, documentation improvements) are usually backported by core dev members. It is also appreciated to make such a backport PR by any contributors, though, so that the overall development proceeds more smoothly!

Issues and Pull Requests

In this section, we explain how to file issues and send pull requests (PRs).

Issue/PR Labels

Issues and PRs are labeled by the following tags:

• Bug: bug reports (issues) and bug fixes (PRs)
• Enhancement: implementation improvements without breaking the interface
• Feature: feature requests (issues) and their implementations (PRs)
• NoCompat: disrupts backward compatibility
• Test: test fixes and updates
• Document: document fixes and improvements
• Example: fixes and improvements on the examples
• Install: fixes installation script
• Contribution-Welcome: issues that we request for contribution (only issues are categorized to this)
• Other: other issues and PRs

Multiple tags might be labeled to one issue/PR. Note that revision releases cannot include PRs in Feature and NoCompat categories.

How to File an Issue

On registering an issue, write precise explanations on how you want CuPy to be. Bug reports must include necessary and sufficient conditions to reproduce the bugs. Feature requests must include what you want to do (and why you want to do, if needed) with CuPy. You can contain your thoughts on how to realize it into the feature requests, though what part is most important for discussions.

Warning

If you have a question on usages of CuPy, it is highly recommended to send a post to CuPy User Group instead of the issue tracker. The issue tracker is not a place to share knowledge on practices. We may suggest these places and immediately close how-to question issues.

How to Send a Pull Request

If you can write code to fix an issue, we encourage to send a PR.

First of all, before starting to write any code, do not forget to confirm the following points.

• Read through the Coding Guidelines and Unit Testing.
• Check the appropriate branch that you should send the PR following Git Branches. If you do not have any idea about selecting a branch, please choose the master branch.

In particular, check the branch before writing any code. The current source tree of the chosen branch is the starting point of your change.

After writing your code (including unit tests and hopefully documentations!), send a PR on GitHub. You have to write a precise explanation of what and how you fix; it is the first documentation of your code that developers read, which is a very important part of your PR.

Once you send a PR, it is automatically tested on Travis CI for Linux and Mac OS X, and on AppVeyor for Windows. Your PR needs to pass at least the test for Linux on Travis CI. After the automatic test passes, some of the core developers will start reviewing your code. Note that this automatic PR test only includes CPU tests.

Note

We are also running continuous integration with GPU tests for the master branch and the versioned branch of the latest major version. Since this service is currently running on our internal server, we do not use it for automatic PR tests to keep the server secure.

If you are planning to add a new feature or modify existing APIs, it is recommended to open an issue and discuss the design first. The design discussion needs lower cost for the core developers than code review. Following the consequences of the discussions, you can send a PR that is smoothly reviewed in a shorter time.

Even if your code is not complete, you can send a pull request as a work-in-progress PR by putting the [WIP] prefix to the PR title. If you write a precise explanation about the PR, core developers and other contributors can join the discussion about how to proceed the PR. WIP PR is also useful to have discussions based on a concrete code.

Coding Guidelines

Note

Coding guidelines are updated at v5.0. Those who have contributed to older versions should read the guidelines again.

We use PEP8 and a part of OpenStack Style Guidelines related to general coding style as our basic style guidelines.

You can use autopep8 and flake8 commands to check your code.

In order to avoid confusion from using different tool versions, we pin the versions of those tools. Install them with the following command (from within the top directory of CuPy repository):

$ pip install -e '.[stylecheck]'

And check your code with:

$ autopep8 path/to/your/code.py
$ flake8 path/to/your/code.py

To check Cython code, use .flake8.cython configuration file:

$ flake8 --config=.flake8.cython path/to/your/cython/code.pyx

The autopep8 supports automatically correct Python code to conform to the PEP 8 style guide:

$ autopep8 --in-place path/to/your/code.py

The flake8 command lets you know the part of your code not obeying our style guidelines. Before sending a pull request, be sure to check that your code passes the flake8 checking.

Note that flake8 command is not perfect. It does not check some of the style guidelines. Here is a (not-complete) list of the rules that flake8 cannot check.

• Relative imports are prohibited. [H304]
• Importing non-module symbols is prohibited.
• Import statements must be organized into three parts: standard libraries, third-party libraries, and internal imports. [H306]

In addition, we restrict the usage of shortcut symbols in our code base. They are symbols imported by packages and sub-packages of cupy. For example, cupy.cuda.Device is a shortcut of cupy.cuda.device.Device. It is not allowed to use such shortcuts in the ``cupy`` library implementation. Note that you can still use them in tests and examples directories.

Once you send a pull request, your coding style is automatically checked by Travis-CI. The reviewing process starts after the check passes.

The CuPy is designed based on NumPy’s API design. CuPy’s source code and documents contain the original NumPy ones. Please note the followings when writing the document.

• In order to identify overlapping parts, it is preferable to add some remarks that this document is just copied or altered from the original one. It is also preferable to briefly explain the specification of the function in a short paragraph, and refer to the corresponding function in NumPy so that users can read the detailed document. However, it is possible to include a complete copy of the document with such a remark if users cannot summarize in such a way.
• If a function in CuPy only implements a limited amount of features in the original one, users should explicitly describe only what is implemented in the document.

For changes that modify or add new Cython files, please make sure the pointer types follow these guidelines (#1913).

• Pointers should be void* if only used within Cython, or intptr_t if exposed to the Python space.
• Memory sizes should be size_t.
• Memory offsets should be ptrdiff_t.

Note

We are incrementally enforcing the above rules, so some existing code may not follow the above guidelines, but please ensure all new contributions do.

Unit Testing

Testing is one of the most important part of your code. You must write test cases and verify your implementation by following our testing guide.

Note that we are using pytest and mock package for testing, so install them before writing your code:

$ pip install pytest mock

How to Run Tests

In order to run unit tests at the repository root, you first have to build Cython files in place by running the following command:

$ pip install -e .

Note

When you modify *.pxd files, before running pip install -e ., you must clean *.cpp and *.so files once with the following command, because Cython does not automatically rebuild those files nicely:

$ git clean -fdx

Note

It’s not officially supported, but you can use ccache to reduce compilation time. On Ubuntu 16.04, you can set up as follows:

$ sudo apt-get install ccache
$ export PATH=/usr/lib/ccache:$PATH

See ccache for details.

If you want to use ccache for nvcc, please install ccache v3.3 or later. You also need to set environment variable NVCC='ccache nvcc'.

Once Cython modules are built, you can run unit tests by running the following command at the repository root:

$ python -m pytest

CUDA must be installed to run unit tests.

Some GPU tests require cuDNN to run. In order to skip unit tests that require cuDNN, specify -m='not cudnn' option:

$ python -m pytest path/to/your/test.py -m='not cudnn'

Some GPU tests involve multiple GPUs. If you want to run GPU tests with insufficient number of GPUs, specify the number of available GPUs to CUPY_TEST_GPU_LIMIT. For example, if you have only one GPU, launch pytest by the following command to skip multi-GPU tests:

$ export CUPY_TEST_GPU_LIMIT=1
$ python -m pytest path/to/gpu/test.py

Following this naming convention, you can run all the tests by running the following command at the repository root:

$ python -m pytest

Or you can also specify a root directory to search test scripts from:

$ python -m pytest tests/cupy_tests     # to just run tests of CuPy
$ python -m pytest tests/install_tests  # to just run tests of installation modules

If you modify the code related to existing unit tests, you must run appropriate commands.

Test File and Directory Naming Conventions

Tests are put into the tests/cupy_tests directory. In order to enable test runner to find test scripts correctly, we are using special naming convention for the test subdirectories and the test scripts.

• The name of each subdirectory of tests must end with the _tests suffix.
• The name of each test script must start with the test_ prefix.

When we write a test for a module, we use the appropriate path and file name for the test script whose correspondence to the tested module is clear. For example, if you want to write a test for a module cupy.x.y.z, the test script must be located at tests/cupy_tests/x_tests/y_tests/test_z.py.

How to Write Tests

There are many examples of unit tests under the tests directory, so reading some of them is a good and recommended way to learn how to write tests for CuPy. They simply use the unittest package of the standard library, while some tests are using utilities from cupy.testing.

In addition to the Coding Guidelines mentioned above, the following rules are applied to the test code:

• All test classes must inherit from unittest.TestCase.

• Use unittest features to write tests, except for the following cases:

  • Use assert statement instead of self.assert* methods (e.g., write assert x == 1 instead of self.assertEqual(x, 1)).
  • Use with pytest.raises(...): instead of with self.assertRaises(...):.

Note

We are incrementally applying the above style. Some existing tests may be using the old style (self.assertRaises, etc.), but all newly written tests should follow the above style.

Even if your patch includes GPU-related code, your tests should not fail without GPU capability. Test functions that require CUDA must be tagged by the cupy.testing.attr.gpu:

import unittest
from cupy.testing import attr

class TestMyFunc(unittest.TestCase):
    ...

    @attr.gpu
    def test_my_gpu_func(self):
        ...

The functions tagged by the gpu decorator are skipped if CUPY_TEST_GPU_LIMIT=0 environment variable is set. We also have the cupy.testing.attr.cudnn decorator to let pytest know that the test depends on cuDNN. The test functions decorated by cudnn are skipped if -m='not cudnn' is given.

The test functions decorated by gpu must not depend on multiple GPUs. In order to write tests for multiple GPUs, use cupy.testing.attr.multi_gpu() decorators instead:

import unittest
from cupy.testing import attr

class TestMyFunc(unittest.TestCase):
    ...

    @attr.multi_gpu(2)  # specify the number of required GPUs here
    def test_my_two_gpu_func(self):
        ...

If your test requires too much time, add cupy.testing.attr.slow decorator. The test functions decorated by slow are skipped if -m='not slow' is given:

import unittest
from cupy.testing import attr

class TestMyFunc(unittest.TestCase):
    ...

    @attr.slow
    def test_my_slow_func(self):
        ...

Note

If you want to specify more than two attributes, use and operator like -m='not cudnn and not slow'. See detail in the document of pytest.

Once you send a pull request, Travis-CI automatically checks if your code meets our coding guidelines described above. Since Travis-CI does not support CUDA, we cannot run unit tests automatically. The reviewing process starts after the automatic check passes. Note that reviewers will test your code without the option to check CUDA-related code.

Note

Some of numerically unstable tests might cause errors irrelevant to your changes. In such a case, we ignore the failures and go on to the review process, so do not worry about it!

Documentation

When adding a new feature to the framework, you also need to document it in the reference.

Note

If you are unsure about how to fix the documentation, you can submit a pull request without doing so. Reviewers will help you fix the documentation appropriately.

The documentation source is stored under docs directory and written in reStructuredText format.

To build the documentation, you need to install Sphinx:

$ pip install sphinx sphinx_rtd_theme

Then you can build the documentation in HTML format locally:

$ cd docs
$ make html

HTML files are generated under build/html directory. Open index.html with the browser and see if it is rendered as expected.

Note

Docstrings (documentation comments in the source code) are collected from the installed CuPy module. If you modified docstrings, make sure to install the module (e.g., using pip install -e .) before building the documentation.

Installation Guide

Recommended Environments

We recommend the following Linux distributions.

• Ubuntu 16.04 / 18.04 LTS (64-bit)
• CentOS 7 (64-bit)

Note

We are automatically testing CuPy on all the recommended environments above. We cannot guarantee that CuPy works on other environments including Windows and macOS, even if CuPy may seem to be running correctly.

Requirements

You need to have the following components to use CuPy.

• NVIDIA CUDA GPU

  • Compute Capability of the GPU must be at least 3.0.

• CUDA Toolkit

  • Supported Versions: 8.0, 9.0, 9.1, 9.2, 10.0, 10.1 and 10.2
  • If you have multiple versions of CUDA Toolkit installed, CuPy will choose one of the CUDA installations automatically. See Working with Custom CUDA Installation for details.

• Python

  • Supported Versions: 3.5.1+, 3.6.0+, 3.7.0+ and 3.8.0+.

• NumPy

  • Supported Versions: 1.9, 1.10, 1.11, 1.12, 1.13, 1.14, 1.15, 1.16, 1.17 and 1.18.
  • NumPy will be installed automatically during the installation of CuPy.

Before installing CuPy, we recommend you to upgrade setuptools and pip:

$ pip install -U setuptools pip

Note

On Windows, CuPy only supports Python 3.6.0 or later.

Note

Python 2 is not supported in CuPy v7.x releases. Please consider migrating Python 3 or use CuPy v6.x, which is the last version that supports Python 2.

Optional Libraries

Some features in CuPy will only be enabled if the corresponding libraries are installed.

• cuDNN (library to accelerate deep neural network computations)

  • Supported Versions: v5, v5.1, v6, v7, v7.1, v7.2, v7.3, v7.4 and v7.5.

• NCCL (library to perform collective multi-GPU / multi-node computations)

  • Supported Versions: v1.3.4, v2, v2.1, v2.2, v2.3 and v2.4.

• cuTENSOR (library for high-performance tensor operations)

  • Supported Versions: v1.0.0 (experimental)

Install CuPy

Wheels (precompiled binary packages) are available for Linux (Python 2.7 or later) and Windows (Python 3.6 or later). Package names are different depending on the CUDA version you have installed on your host.

(For CUDA 8.0)
$ pip install cupy-cuda80

(For CUDA 9.0)
$ pip install cupy-cuda90

(For CUDA 9.1)
$ pip install cupy-cuda91

(For CUDA 9.2)
$ pip install cupy-cuda92

(For CUDA 10.0)
$ pip install cupy-cuda100

(For CUDA 10.1)
$ pip install cupy-cuda101

(For CUDA 10.2)
$ pip install cupy-cuda102

Note

The latest version of cuDNN and NCCL libraries are included in these wheels. You don’t have to install them manually.

When using wheels, please be careful not to install multiple CuPy packages at the same time. Any of these packages and cupy package (source installation) conflict with each other. Please make sure that only one CuPy package (cupy or cupy-cudaXX where XX is a CUDA version) is installed:

$ pip freeze | grep cupy

Install CuPy from conda-forge

Conda/Anaconda is a cross-platform package management solution widely used in scientific computing and other fields. The above pip install instruction is compatible with conda environments. Alternatively, for Linux 64 systems once the CUDA driver is correctly set up, you can install CuPy from the conda-forge channel:

$ conda install -c conda-forge cupy

and conda will install pre-built CuPy and most of the optional dependencies for you, including CUDA runtime libraries (cudatoolkit), NCCL, and cuDNN. It is not necessary to install CUDA Toolkit in advance. If you need to enforce the installation of a particular CUDA version (say 10.0) for driver compatibility, you can do:

$ conda install -c conda-forge cupy cudatoolkit=10.0

Note

Currently cuTENSOR is not yet available on conda-forge.

Note

If you encounter any problem with CuPy from conda-forge, please feel free to report to cupy-feedstock, and we will help investigate if it is just a packaging issue in conda-forge’s recipe or a real issue in CuPy.

Note

If you did not install CUDA Toolkit yourselves, the nvcc compiler might not be available. The cudatoolkit package from Anaconda does not have nvcc included.

Install CuPy from Source

It is recommended to use wheels whenever possible. However, if wheels cannot meet your requirements (e.g., you are running non-Linux environment or want to use a version of CUDA / cuDNN / NCCL not supported by wheels), you can also build CuPy from source.

When installing from source, C++ compiler such as g++ is required. You need to install it before installing CuPy. This is typical installation method for each platform:

# Ubuntu 16.04
$ apt-get install g++

# CentOS 7
$ yum install gcc-c++

Note

When installing CuPy from source, features provided by optional libraries (cuDNN and NCCL) will be disabled if these libraries are not available at the time of installation. See Installing cuDNN and NCCL for the instructions.

Note

If you upgrade or downgrade the version of CUDA Toolkit, cuDNN or NCCL, you may need to reinstall CuPy. See Reinstall CuPy for details.

Using pip

You can install CuPy package via pip.

$ pip install cupy

Using Tarball

The tarball of the source tree is available via pip download cupy or from the release notes page. You can install CuPy from the tarball:

$ pip install cupy-x.x.x.tar.gz

You can also install the development version of CuPy from a cloned Git repository:

$ git clone https://github.com/cupy/cupy.git
$ cd cupy
$ pip install .

If you are using source tree downloaded from GitHub, you need to install Cython 0.28.0 or later (pip install cython).

Uninstall CuPy

Use pip to uninstall CuPy:

$ pip uninstall cupy

Note

When you upgrade Chainer, pip sometimes installs the new version without removing the old one in site-packages. In this case, pip uninstall only removes the latest one. To ensure that CuPy is completely removed, run the above command repeatedly until pip returns an error.

Note

If you are using a wheel, cupy shall be replaced with cupy-cudaXX (where XX is a CUDA version number).

Note

If CuPy is installed via conda, please do conda uninstall cupy instead.

Upgrade CuPy

Just use pip install with -U option:

$ pip install -U cupy

Note

If you are using a wheel, cupy shall be replaced with cupy-cudaXX (where XX is a CUDA version number).

Reinstall CuPy

If you want to reinstall CuPy, please uninstall CuPy and then install it. When reinstalling CuPy, we recommend to use --no-cache-dir option as pip caches the previously built binaries:

$ pip uninstall cupy
$ pip install cupy --no-cache-dir

Note

If you are using a wheel, cupy shall be replaced with cupy-cudaXX (where XX is a CUDA version number).

Run CuPy with Docker

We are providing the official Docker image. Use nvidia-docker command to run CuPy image with GPU. You can login to the environment with bash, and run the Python interpreter:

$ nvidia-docker run -it cupy/cupy /bin/bash

Or run the interpreter directly:

$ nvidia-docker run -it cupy/cupy /usr/bin/python

FAQ

Warning message “cuDNN is not enabled” appears when using Chainer

You failed to build CuPy with cuDNN. If you don’t need cuDNN, ignore this message. Otherwise, retry to install CuPy with cuDNN.

See Installing cuDNN and NCCL and pip fails to install CuPy for details.

pip fails to install CuPy

Please make sure that you are using the latest setuptools and pip:

$ pip install -U setuptools pip

Use -vvvv option with pip command. This will display all logs of installation:

$ pip install cupy -vvvv

If you are using sudo to install CuPy, note that sudo command does not propagate environment variables. If you need to pass environment variable (e.g., CUDA_PATH), you need to specify them inside sudo like this:

$ sudo CUDA_PATH=/opt/nvidia/cuda pip install cupy

If you are using certain versions of conda, it may fail to build CuPy with error g++: error: unrecognized command line option ‘-R’. This is due to a bug in conda (see conda/conda#6030 for details). If you encounter this problem, please upgrade your conda.

Installing cuDNN and NCCL

We recommend installing cuDNN and NCCL using binary packages (i.e., using apt or yum) provided by NVIDIA.

If you want to install tar-gz version of cuDNN and NCCL, we recommend you to install it under CUDA directory. For example, if you are using Ubuntu, copy *.h files to include directory and *.so* files to lib64 directory:

$ cp /path/to/cudnn.h $CUDA_PATH/include
$ cp /path/to/libcudnn.so* $CUDA_PATH/lib64

The destination directories depend on your environment.

If you want to use cuDNN or NCCL installed in another directory, please use CFLAGS, LDFLAGS and LD_LIBRARY_PATH environment variables before installing CuPy:

export CFLAGS=-I/path/to/cudnn/include
export LDFLAGS=-L/path/to/cudnn/lib
export LD_LIBRARY_PATH=/path/to/cudnn/lib:$LD_LIBRARY_PATH

Note

Use full paths for the environment variables. distutils that is used in the setup script does not expand the home directory mark ~.

Working with Custom CUDA Installation

If you have installed CUDA on the non-default directory or have multiple CUDA versions installed, you may need to manually specify the CUDA installation directory to be used by CuPy.

CuPy uses the first CUDA installation directory found by the following order.

1. CUDA_PATH environment variable.
2. The parent directory of nvcc command. CuPy looks for nvcc command in each directory set in PATH environment variable.
3. /usr/local/cuda

For example, you can tell CuPy to use non-default CUDA directory by CUDA_PATH environment variable:

$ CUDA_PATH=/opt/nvidia/cuda pip install cupy

Note

CUDA installation discovery is also performed at runtime using the rule above. Depending on your system configuration, you may also need to set LD_LIBRARY_PATH environment variable to $CUDA_PATH/lib64 at runtime.

Using custom nvcc command during installation

If you want to use a custom nvcc compiler (for example, to use ccache) to build CuPy, please set NVCC environment variables before installing CuPy:

export NVCC='ccache nvcc'

Note

During runtime, you don’t need to set this environment variable since CuPy doesn’t use the nvcc command.

Installation for Developers

If you are hacking CuPy source code, we recommend you to use pip with -e option for editable mode:

$ cd /path/to/cupy/source
$ pip install -e .

Please note that even with -e, you will have to rerun pip install -e . to regenerate C++ sources using Cython if you modified Cython source files (e.g., *.pyx files).

CuPy always raises cupy.cuda.compiler.CompileException

If CuPy does not work at all with CompileException, it is possible that CuPy cannot detect CUDA installed on your system correctly. The followings are error messages commonly observed in such cases.

• nvrtc: error: failed to load builtins
• catastrophic error: cannot open source file "cuda_fp16.h"
• error: cannot overload functions distinguished by return type alone
• error: identifier "__half_raw" is undefined

Please try setting LD_LIBRARY_PATH and CUDA_PATH environment variable. For example, if you have CUDA installed at /usr/local/cuda-9.0:

export CUDA_PATH=/usr/local/cuda-9.0
export LD_LIBRARY_PATH=$CUDA_PATH/lib64:$LD_LIBRARY_PATH

Also see Working with Custom CUDA Installation.

[Experimental] Installation Guide for ROCm environemt

This is an experimental feature. We recommend only for advanced users to use this.

Recommended Environments

We recommend the following Linux distributions.

• Ubuntu 16.04 / 18.04 LTS (64-bit)

Requirements

You need to have the following components to use CuPy.

• GPU supported by ROCm (AMD GPUs or NVIDIA GPUs)

• ROCm

  • Supported Versions: ROCm 2.6+.

• Python
• NumPy

And please install ROCm libraries.

$ sudo apt install hipblas hipsparse rocrand rocthrust

Before installing CuPy, we recommend you to upgrade setuptools and pip:

$ pip install -U setuptools pip

Install CuPy from Source

It is recommended to use wheels whenever possible. However, there is currently no wheels for the ROCm environment, so you have to build it from source.

When installing from source, C++ compiler such as g++ is required. You need to install it before installing CuPy. This is typical installation method for each platform:

# Ubuntu 16.04
$ apt-get install g++

Note

If you want to upgrade or downgrade the version of ROCm, you may need to reinstall CuPy after that. See Reinstall CuPy for details.

Using pip

You can install CuPy package via pip. It builds CuPy from source.

$ export HCC_AMDGPU_TARGET=gfx900  # This value should be changed based on your GPU
$ export __HIP_PLATFORM_HCC__
$ export CUPY_INSTALL_USE_HIP=1
$ pip install cupy

Using Tarball

The tarball of the source tree is available via pip download cupy or from the release notes page. You can install CuPy from the tarball:

$ pip install cupy-x.x.x.tar.gz

You can also install the development version of CuPy from a cloned Git repository:

$ git clone https://github.com/cupy/cupy.git
$ cd cupy
$ export HCC_AMDGPU_TARGET=gfx900  # This value should be changed based on your GPU
$ export __HIP_PLATFORM_HCC__
$ export CUPY_INSTALL_USE_HIP=1
$ pip install .

If you are using the source tree downloaded from GitHub, you need to install Cython 0.28.0 or later (pip install cython).

Uninstall CuPy

Use pip to uninstall CuPy:

$ pip uninstall cupy

Note

When you upgrade Chainer, pip sometimes installs the new version without removing the old one in site-packages. In this case, pip uninstall only removes the latest one. To ensure that CuPy is completely removed, run the above command repeatedly until pip returns an error.

Upgrade CuPy

Just use pip install with -U option:

$ export HCC_AMDGPU_TARGET=gfx900  # This value should be changed based on your GPU
$ export __HIP_PLATFORM_HCC__
$ export CUPY_INSTALL_USE_HIP=1
$ pip install -U cupy

Reinstall CuPy

If you want to reinstall CuPy, please uninstall CuPy first, and then install again. When reinstalling CuPy, we recommend to use --no-cache-dir option as pip caches the previously built binaries:

$ pip uninstall cupy
$ export HCC_AMDGPU_TARGET=gfx900  # This value should be changed based on your GPU
$ export __HIP_PLATFORM_HCC__
$ export CUPY_INSTALL_USE_HIP=1
$ pip install cupy --no-cache-dir

FAQ

pip fails to install CuPy

Please make sure that you are using the latest setuptools and pip:

$ pip install -U setuptools pip

Use -vvvv option with pip command to investigate the details of errors. This will display all logs of installation:

$ pip install cupy -vvvv

If you are using sudo to install CuPy, note that sudo command does not propagate environment variables. If you need to pass environment variable (e.g., ROCM_HOME), you need to specify them inside sudo like this:

$ sudo ROCM_HOME=/opt/rocm pip install cupy

If you are using certain versions of conda, it may fail to build CuPy with error g++: error: unrecognized command line option ‘-R’. This is due to a bug in conda (see conda/conda#6030 for details). If you encounter this problem, please downgrade or upgrade it.

Upgrade Guide

This is a list of changes introduced in each release that users should be aware of when migrating from older versions. Most changes are carefully designed not to break existing code; however changes that may possibly break them are highlighted with a box.

CuPy v7

Dropping Support of Python 2.7 and 3.4

Starting from CuPy v7, Python 2.7 and 3.4 are no longer supported as it reaches its end-of-life (EOL) in January 2020 (2.7) and March 2019 (3.4). Python 3.5.1 is the minimum Python version supported by CuPy v7. Please upgrade the Python version if you are using affected versions of Python to any later versions listed under Installation.

CuPy v6

Binary Packages Ignore LD_LIBRARY_PATH

Prior to CuPy v6, LD_LIBRARY_PATH environment variable can be used to override cuDNN / NCCL libraries bundled in the binary distribution (also known as wheels). In CuPy v6, LD_LIBRARY_PATH will be ignored during discovery of cuDNN / NCCL; CuPy binary distributions always use libraries that comes with the package to avoid errors caused by unexpected override.

CuPy v5

cupyx.scipy Namespace

cupyx.scipy namespace has been introduced to provide CUDA-enabled SciPy functions. cupy.sparse module has been renamed to cupyx.scipy.sparse; cupy.sparse will be kept as an alias for backward compatibility.

Dropped Support for CUDA 7.0 / 7.5

CuPy v5 no longer supports CUDA 7.0 / 7.5.

Update of Docker Images

CuPy official Docker images (see Installation Guide for details) are now updated to use CUDA 9.2 and cuDNN 7.

To use these images, you may need to upgrade the NVIDIA driver on your host. See Requirements of nvidia-docker for details.

CuPy v4

Note

The version number has been bumped from v2 to v4 to align with the versioning of Chainer. Therefore, CuPy v3 does not exist.

Default Memory Pool

Prior to CuPy v4, memory pool was only enabled by default when CuPy is used with Chainer. In CuPy v4, memory pool is now enabled by default, even when you use CuPy without Chainer. The memory pool significantly improves the performance by mitigating the overhead of memory allocation and CPU/GPU synchronization.

Attention

When you monitor GPU memory usage (e.g., using nvidia-smi), you may notice that GPU memory not being freed even after the array instance become out of scope. This is expected behavior, as the default memory pool “caches” the allocated memory blocks.

To access the default memory pool instance, use get_default_memory_pool() and get_default_pinned_memory_pool(). You can access the statistics and free all unused memory blocks “cached” in the memory pool.

import cupy
a = cupy.ndarray(100, dtype=cupy.float32)
mempool = cupy.get_default_memory_pool()

# For performance, the size of actual allocation may become larger than the requested array size.
print(mempool.used_bytes())   # 512
print(mempool.total_bytes())  # 512

# Even if the array goes out of scope, its memory block is kept in the pool.
a = None
print(mempool.used_bytes())   # 0
print(mempool.total_bytes())  # 512

# You can clear the memory block by calling `free_all_blocks`.
mempool.free_all_blocks()
print(mempool.used_bytes())   # 0
print(mempool.total_bytes())  # 0

You can even disable the default memory pool by the code below. Be sure to do this before any other CuPy operations.

import cupy
cupy.cuda.set_allocator(None)
cupy.cuda.set_pinned_memory_allocator(None)

Compute Capability

CuPy v4 now requires NVIDIA GPU with Compute Capability 3.0 or larger. See the List of CUDA GPUs to check if your GPU supports Compute Capability 3.0.

CUDA Stream

As CUDA Stream is fully supported in CuPy v4, cupy.cuda.RandomState.set_stream, the function to change the stream used by the random number generator, has been removed. Please use cupy.cuda.Stream.use() instead.

See the discussion in #306 for more details.

cupyx Namespace

cupyx namespace has been introduced to provide features specific to CuPy (i.e., features not provided in NumPy) while avoiding collision in future. See CuPy-specific Functions for the list of such functions.

For this rule, cupy.scatter_add() has been moved to cupyx.scatter_add(). cupy.scatter_add() is still available as an alias, but it is encouraged to use cupyx.scatter_add() instead.

Update of Docker Images

CuPy official Docker images (see Installation Guide for details) are now updated to use CUDA 8.0 and cuDNN 6.0. This change was introduced because CUDA 7.5 does not support NVIDIA Pascal GPUs.

To use these images, you may need to upgrade the NVIDIA driver on your host. See Requirements of nvidia-docker for details.

CuPy v2

Changed Behavior of count_nonzero Function

For performance reasons, cupy.count_nonzero() has been changed to return zero-dimensional ndarray instead of int when axis=None. See the discussion in #154 for more details.

License

Copyright (c) 2015 Preferred Infrastructure, Inc.

Copyright (c) 2015 Preferred Networks, Inc.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

NumPy

The CuPy is designed based on NumPy’s API. CuPy’s source code and documents contain the original NumPy ones.

Copyright (c) 2005-2016, NumPy Developers.

All rights reserved.

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

• Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
• Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
• Neither the name of the NumPy Developers nor the names of any contributors may be used to endorse or promote products derived from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

SciPy

The CuPy is designed based on SciPy’s API. CuPy’s source code and documents contain the original SciPy ones.

Copyright (c) 2001, 2002 Enthought, Inc.

All rights reserved.

Copyright (c) 2003-2016 SciPy Developers.

All rights reserved.

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
3. Neither the name of Enthought nor the names of the SciPy Developers may be used to endorse or promote products derived from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.