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 and cuSOLVERMulti-dimensional fast Fourier transform (FFT), accelerated by cuFFT
Reduction along axes (
sum
,max
,argmax
, etc.)
CuPy additionally supports a subset of SciPy features:
Sparse matrices and sparse linear algebra, powered by cuSPARSE.
Fast Fourier transform (FFT)
CuPy also includes the following features for performance:
User-defined elementwise CUDA kernels
User-defined reduction CUDA kernels
Just-in-time compiler converting Python functions to CUDA kernels
Fusing CUDA kernels to optimize user-defined calculation
Customizable memory allocator and memory pool
cuDNN utilities
Full coverage of NCCL APIs
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.