Meta tags:
description=Run NumPy/SciPy code on GPU. Supports NVIDIA CUDA and AMD ROCm.;

Headings (most frequently used words):

by, with, easy, to, watch, videos, developed, supported, high, performance, gpu, highly, compatible, numpy, scipy, install, write, custom, kernel,

Text of the page (most frequently used words):
cupy (21), and (10), the (9), scipy (8), for (8), with (7), numpy (7), gpu (7), float32 (6), cuda (6), code (5), dtype (5), array (5), install (5), kernel (4), you (4), can (4), make (4), pip (4), arange (3), read (3), use (3), more (3), #compatible (3), python (3), preferred (2), inc (2), videos (2), reshape (2), custom (2), your (2), binary (2), tutorial (2), also (2), raw (2), kernels (2), easy (2), rocm (2), from (2), source (2), highly (2), most (2), shows (2), library (2), accelerated (2), computing (2), operations (2), blog (2), github (2), menu (2), infrastructure, design, styleshout, conduct, privacy, policy, networks, supported, developed, 2025, tacoma, pybay, 2019, san, francisco, watch, elementwisekernel, else, my_kernel, easily, want, run, faster, requiring, only, small, snippet, automatically, wraps, compiles, compiled, binaries, are, cached, reused, subsequent, runs, please, via, modules, user, defined, write, cuda12x, cuda13x, amd, easiest, way, provides, wheels, precompiled, packages, linux, windows, details, installed, conda, forge, installation, guide, import, astype, sum, axis, interface, cases, used, drop, replacement, all, need, just, replace, useful, learn, first, steps, supports, various, methods, indexing, data, types, broadcasting, list, apis, their, corresponding, implementations, this, comparison, table, basics, cupyx, open, utilizes, toolkit, libraries, including, cublas, curand, cusolver, cusparse, cufft, cudnn, nccl, full, architecture, figure, speedup, over, perform, well, using, out, box, speeds, some, than, 100x, original, benchmark, article, rapids, medium, single, speedups, high, performance, api, reference, get, started, about, gitter, linkedin, docs, features, hide, show,

Text of the page (random words):
cupy numpy scipy for gpu show menu hide menu features videos blog docs github x linkedin gitter about us numpy scipy compatible array library for gpu accelerated computing with python get started api reference github high performance with gpu cupy is an open source array library for gpu accelerated computing with python cupy utilizes cuda toolkit libraries including cublas curand cusolver cusparse cufft cudnn and nccl to make full use of the gpu architecture the figure shows cupy speedup over numpy most operations perform well on a gpu using cupy out of the box cupy speeds up some operations more than 100x read the original benchmark article single gpu cupy speedups on the rapids ai medium blog highly compatible with numpy scipy cupy s interface is highly compatible with numpy and scipy in most cases it can be used as a drop in replacement all you need to do is just replace numpy and scipy with cupy and cupyx scipy in your python code the basics of cupy tutorial is useful to learn first steps with cupy cupy supports various methods indexing data types broadcasting and more this comparison table shows a list of numpy scipy apis and their corresponding cupy implementations import cupy as cp x cp arange 6 reshape 2 3 astype f x array 0 1 2 3 4 5 dtype float32 x sum axis 1 array 3 12 dtype float32 easy to install the easiest way to install cupy is to use pip cupy provides wheels precompiled binary packages for linux and windows read the installation guide for more details cupy can also be installed from conda forge or from source code for cuda 12 x pip install cupy cuda12x for cuda 13 x pip install cupy cuda13x for amd rocm 7 0 pip install cupy rocm 7 0 easy to write a custom kernel you can easily make a custom cuda kernel if you want to make your code run faster requiring only a small code snippet of c cupy automatically wraps and compiles it to make a cuda binary compiled binaries are cached and reused in subsequent runs please read the user defined kernels tutorial and you can also use raw cuda kernels via raw modules x cp arange 6 dtype f reshape 2 3 y cp arange 3 dtype f kernel cp elementwisekernel float32 x float32 y float32 z if x 2 y z x y else z x y my_kernel kernel x y array 0 2 4 0 4 10 dtype float32 watch videos pybay 2019 san francisco scipy 2025 tacoma developed by supported by preferred networks inc preferred infrastructure inc privacy policy code of conduct design by styleshout