GeezerVison

This article has multiple issues. Please help improve it or discuss these issues on the talk page. (Learn how and when to remove these messages) This article contains instructions or advice. Wikipedia is not a guidebook; please help rewrite such content to be encyclopedic or move it to Wikiversity, Wikibooks, or Wikivoyage. (February 2024) (Learn how and when to remove this message) This article contains a pro and con list. Please help rewrite it into consolidated sections based on topics. (February 2024) (Learn how and when to remove this message)

CUDA
Original authors	Ian Buck John Nickolls
Developer	Nvidia
Release	February 16, 2007 (2007-02-16)[1]
Stable release	13.4.1[2] / 9 September 2026 (9 September 2026)
Written in	C
Operating system	Windows, Linux
Platform	Supported GPUs
Type	GPGPU
License	Proprietary
Website	developer.nvidia.com/cuda-zone

CUDA (Compute Unified Device Architecture) is a proprietary^[3] parallel computing platform and application programming interface (API) developed by Nvidia that allows software to use certain types of graphics processing units (GPUs) for accelerated general-purpose processing, significantly broadening their utility in artificial intelligence, scientific and high-performance computing. CUDA was created in 2004 and was officially released in 2007.^[4] When introduced, the name was an acronym for Compute Unified Device Architecture,^[5] but Nvidia later dropped the initial meaning of the acronym and now rarely expands it.^[6]

CUDA is both a software layer that manages data, giving direct access to the GPU and CPU as necessary, and a library of APIs that enable parallel computation.^[7][8] In addition to drivers and runtime kernels, the CUDA platform includes compilers, libraries and developer tools to help programmers.

CUDA is written in the C programming language, but is designed to work with other programming languages including C++, Fortran, Python and Julia. This accessibility makes it easier for specialists in parallel programming to use GPU resources, in contrast to prior APIs like Direct3D and OpenGL, which require advanced skills in graphics programming.^[9] CUDA-powered GPUs support programming frameworks such as OpenMP, OpenACC and OpenCL.^[10][7]

The graphics processing unit (GPU) is a specialized processor. It addresses the demands of real-time high-resolution, compute-intensive tasks such as 3D graphics. By 2012, GPUs had evolved into highly parallel multi-core systems allowing efficient manipulation of large blocks of data. This design is more effective than general-purpose central processing unit (CPUs) for algorithms in situations where processing large blocks of data is done in parallel, such as:

• cryptographic hash functions
• machine learning
• molecular dynamics simulations
• physics engines

CUDA traces to the early 2000s, when Ian Buck, a computer science PhD student at Stanford University, began experimenting with using GPUs for purposes beyond rendering graphics. Buck had become interested in GPUs during his undergraduate studies at Princeton University, initially through video gaming. After graduation, he interned at Nvidia, gaining deeper exposure to GPU architecture. At Stanford, he built an 8K gaming rig using 32 GeForce graphics cards, originally to push the limits of graphics performance in games like Quake and Doom. However, his interests shifted toward exploring the potential of GPUs for general-purpose parallel computing.^[11]

To that end, Buck developed Brook, a programming language designed to enable general-purpose computing on GPUs. His work attracted support from Nvidia and the Defense Advanced Research Projects Agency (DARPA). In 2004, Nvidia hired Buck and paired him with John Nickolls,^[12] then director of architecture for GPU computing. Together, they began transforming Brook into CUDA.^[11] CUDA was officially released in 2007.

CUDA became central to the company's strategy of positioning GPUs as versatile hardware for scientific applications. By 2015, CUDA's development increasingly focused on accelerating machine learning and artificial neural network workloads.^[13]

The following table offers an approximate summary of the CUDA ontology.

The CUDA platform is accessible to software developers through CUDA-accelerated libraries, compiler directives such as OpenACC, and extensions to industry-standard programming languages including C, C++, Fortran and Python. C/C++ programmers can use 'CUDA C/C++', compiled to PTX with nvcc (Nvidia's LLVM-based C/C++ compiler)^[14] or by clang itself.^[15] Fortran programmers can use 'CUDA Fortran', compiled with the PGI CUDA Fortran compiler from The Portland Group. Python programmers can use the cuPyNumeric library to accelerate applications on Nvidia GPUs.

In addition to libraries, compiler directives, CUDA C/C++ and CUDA Fortran, the CUDA platform supports other computational interfaces, including the Khronos Group's OpenCL,^[16] Microsoft's DirectCompute, OpenGL Compute Shader and C++ AMP.^[17] Third party wrappers are also available for Python, Perl, Fortran, Java, Ruby, Lua, Common Lisp, Haskell, R, MATLAB, IDL, Julia, and native support in Mathematica.

In the computer game industry, GPUs are used for graphics rendering, and for game physics calculations (physical effects such as debris, smoke, fire, fluids); examples include PhysX and Bullet. CUDA has also been used to accelerate non-graphical applications in computational biology, cryptography and other fields by an order of magnitude or more.^{[18][19][20][21][22]}

CUDA provides both a low level API (CUDA Driver API, non single-source) and a higher level API (CUDA Runtime API, single-source). The initial CUDA SDK was made public on 15 February 2007, for Microsoft Windows and Linux. Mac OS X support was later added in version 2.0,^[23] which supersedes the beta released February 14, 2008.^[24] CUDA works with all Nvidia GPUs from the G8x series onwards, including GeForce, Quadro and the Tesla line. CUDA is compatible with most standard operating systems.

CUDA 8.0 comes with the following libraries (for compilation & runtime, in alphabetical order):

• cuBLAS – CUDA Basic Linear Algebra Subroutines library
• CUDART – CUDA Runtime library
• cuFFT – CUDA Fast Fourier Transform library
• cuRAND – CUDA Random Number Generation library
• cuSOLVER – CUDA based collection of dense and sparse direct solvers
• cuSPARSE – CUDA Sparse Matrix library
• NPP – NVIDIA Performance Primitives library
• nvGRAPH – NVIDIA Graph Analytics library
• NVML – NVIDIA Management Library
• NVRTC – NVIDIA Runtime Compilation library for CUDA C++

CUDA 8.0 comes with these other software components:

• nView – NVIDIA nView Desktop Management Software
• NVWMI – NVIDIA Enterprise Management Toolkit
• GameWorks PhysX – is a multi-platform game physics engine

CUDA 9.0–9.2 comes with these other components:

• CUTLASS 1.0 – custom linear algebra algorithms,
• NVIDIA Video Decoder was deprecated in CUDA 9.2; it is now available in NVIDIA Video Codec SDK

CUDA 10 comes with these other components:

• nvJPEG – Hybrid (CPU and GPU) JPEG processing

CUDA 11.0–11.8 comes with these other components:^{[25][26][27][28]}

• CUB is new one of more supported C++ libraries
• MIG multi instance GPU support
• nvJPEG2000 – JPEG 2000 encoder and decoder

CUDA has several advantages over traditional general-purpose computation on GPUs (GPGPU) using graphics APIs:

• Scattered reads – code can read from arbitrary addresses in memory
• Unified virtual memory (CUDA 4.0 and above)
• Unified memory (CUDA 6.0 and above)
• Shared memory – CUDA exposes a fast shared memory region that can be shared among threads. This can be used as a user-managed cache, enabling higher bandwidth than is possible using texture lookups.^[29]
• Faster downloads and readbacks to and from the GPU
• Full support for integer and bitwise operations, including integer texture lookups

• Whether for the host computer or the GPU device, all CUDA source code is now processed according to C++ syntax rules.^[30] This was not always the case. Earlier versions of CUDA were based on C syntax rules.^[31] As with the more general case of compiling C code with a C++ compiler, it is therefore possible that old C-style CUDA source code will either fail to compile or will not behave as originally intended.
• Interoperability with rendering languages such as OpenGL is one-way, with OpenGL having access to registered CUDA memory but CUDA not having access to OpenGL memory.
• Copying between host and device memory may incur a performance hit due to system bus bandwidth and latency (this can be partly alleviated with asynchronous memory transfers, handled by the GPU's DMA engine).
• Threads should be running in groups of at least 32 for best performance, with total number of threads numbering in the thousands. Branches in the program code do not affect performance significantly, provided that each of 32 threads takes the same execution path; the SIMD execution model becomes a significant limitation for any inherently divergent task (e.g. traversing a space partitioning data structure during ray tracing).
• No emulation or fallback functionality is available for modern revisions.
• Valid C++ may sometimes be flagged and prevent compilation due to the way the compiler approaches optimization for target GPU device limitations.
• C++ run-time type information (RTTI) and C++-style exception handling are only supported in host code, not in device code.
• In single-precision on first generation CUDA compute capability 1.x devices, denormal numbers are unsupported and are instead flushed to zero, and the precision of both the division and square root operations are slightly lower than IEEE 754-compliant single precision math. Devices that support compute capability 2.0 and above support denormal numbers, and the division and square root operations are IEEE 754 compliant by default. However, users can obtain the prior faster gaming-grade math of compute capability 1.x devices if desired by setting compiler flags to disable accurate divisions and accurate square roots, and enable flushing denormal numbers to zero.^[32]
• Unlike OpenCL, CUDA-enabled GPUs are only available from Nvidia as it is proprietary.^[33][3] Attempts to implement CUDA on other GPUs include:
  • Project Coriander: Converts CUDA C++11 source to OpenCL 1.2 C. A fork of CUDA-on-CL intended to run TensorFlow.^[34][35][36]
  • CU2CL: Convert CUDA 3.2 C++ to OpenCL C.^[37]
  • GPUOpen HIP: A thin abstraction layer on top of CUDA and ROCm intended for AMD and Nvidia GPUs. Has a conversion tool for importing CUDA C++ source. Supports CUDA 4.0 plus C++11 and float16.
  • ZLUDA is a drop-in replacement for CUDA on AMD GPUs and formerly Intel GPUs with near-native performance.^[38] The developer, Andrzej Janik, was separately contracted by both Intel and AMD to develop the software in 2021 and 2022, respectively. However, neither company decided to release it officially due to the lack of a business use case. AMD's contract included a clause that allowed Janik to release his code for AMD independently, allowing him to release the new version that only supports AMD GPUs.^[39]
  • ChipStar can compile and run CUDA/HIP programs on advanced OpenCL 3.0 or Level Zero platforms.^[40]
  • SCALE is a CUDA-compatible programming toolkit for ahead of time compilation of CUDA source code on AMD GPUs, aiming to expand support for other GPUs in the future.^[41]

This example code in C++ loads a texture from an image into an array on the GPU:

void foo()
{
    cudaArray* cu_array = nullptr;

    // Allocate array
    cudaChannelFormatDesc description = cudaCreateChannelDesc<float>();
    cudaMallocArray(&cu_array, &description, width, height);

    // Copy image data to array
    cudaMemcpy2DToArray(cu_array, 0, 0,
                        image, width * sizeof(float),
                        width * sizeof(float), height,
                        cudaMemcpyHostToDevice);

    // Resource description (what the texture points to)
    cudaResourceDesc resDesc = {};
    resDesc.resType           = cudaResourceTypeArray;
    resDesc.res{.array{.array = cu_array}};

    // Texture description (sampling parameters)
    cudaTextureDesc texDesc = {};
    texDesc.addressMode      = {cudaAddressModeClamp, cudaAddressModeClamp};
    texDesc.filterMode       = cudaFilterModePoint;
    texDesc.readMode         = cudaReadModeElementType;
    texDesc.normalizedCoords = 0; // use unnormalized integer coordinates

    // Create the texture object
    cudaTextureObject_t texObj = 0;
    cudaCreateTextureObject(&texObj, &resDesc, &texDesc, nullptr);

    // Run kernel - pass the texture object as a parameter
    dim3 blockDim(16, 16, 1);
    dim3 gridDim((width  + blockDim.x - 1) / blockDim.x,
                 (height + blockDim.y - 1) / blockDim.y, 1);
    kernel<<<gridDim, blockDim>>>(texObj, d_data, height, width);

    // Clean up
    cudaDestroyTextureObject(texObj);
    // cudaFreeArray(cu_array); // when you are done with the array
}

__global__ void kernel(cudaTextureObject_t texObj,
                       float* odata, int height, int width)
{
    int x = blockIdx.x * blockDim.x + threadIdx.x;
    int y = blockIdx.y * blockDim.y + threadIdx.y;

    if (x < width && y < height)
    {
        float c = tex2D<float>(texObj, x, y);
        odata[y * width + x] = c;
    }
}

Below is an example given in Python that computes the product of two arrays on the GPU. The unofficial Python language bindings can be obtained from PyCUDA.^[42]

import numpy
import pycuda.autoinit

from numpy.typing import NDArray, float32
from pycuda.compiler import SourceModule
from pycuda.driver import Function, In, Out

mod: SourceModule = SourceModule(
    """
__global__ void multiply_them(float* dest, float* a, float* b) {
    const int i = threadIdx.x;
    dest[i] = a[i] * b[i];
}
"""
)

multiply_them: Function = mod.get_function("multiply_them")

a: NDArray[float32] = numpy.random.randn(400).astype(numpy.float32)
b: NDArray[float32] = numpy.random.randn(400).astype(numpy.float32)

dest: NDArray[float32] = numpy.zeros_like(a)
multiply_them(Out(dest), In(a), In(b), block=(400, 1, 1))

print(dest - a * b)

Additional Python bindings to simplify matrix multiplication operations can be found in the program pycublas.^[43]

 
import numpy

from pycublas import CUBLASMatrix

A: CUBLASMatrix = CUBLASMatrix(numpy.mat([[1, 2, 3], [4, 5, 6]], numpy.float32))
B: CUBLASMatrix = CUBLASMatrix(numpy.mat([[2, 3], [4, 5], [6, 7]], numpy.float32))
C: CUBLASMatrix = A * B
print(C.np_mat())

while CuPy directly replaces NumPy:^[44]

import cupy

from cupy.typing import NDArray, float64

a: NDArray[float64] = cupy.random.randn(400)
b: NDArray[float64] = cupy.random.randn(400)

dest: NDArray[float64] = cupy.zeros_like(a)

print(dest - a * b)

Note on notation: compute capability X.Y is also written as SMXY or sm_XY (e.g. 10.3 as SM103 or sm_103) in professional Nvidia software and the code Nvidia has contributed to LLVM.^[45]

Below is a table of supported CUDA compute capabilities based on the CUDA SDK version and microarchitecture, listed by code name:

Note: CUDA SDK 10.2 is the last official release for macOS, as support will not be available for macOS in newer releases.

CUDA compute capability by version with associated GPU semiconductors and GPU card models (separated by their various application areas):

* – OEM-only products

Note: A GPU with a higher compute capacity is able to execute PTX code meant for a GPU of a lower range of compute capacities. However, it is possible to compile CUDA code into a form that only works on one family (same "X") of GPUs; if existing code is compiled this way, recompilation will be needed for it to work on a newer GPU.^[45]

^[66]

• Accelerated rendering of 3D graphics
• Accelerated interconversion of video file formats
• Accelerated encryption, decryption and compression
• Bioinformatics, e.g. NGS DNA sequencing BarraCUDA^[123]
• Distributed calculations, such as predicting the native conformation of proteins
• Medical analysis simulations, for example virtual reality based on CT and MRI scan images
• Physical simulations,^[124] particularly in fluid dynamics
• Neural network training in machine learning problems
• Large Language Model inference
• Face recognition
• Volunteer computing projects, such as SETI@home and other projects using BOINC software
• Molecular dynamics
• Mining cryptocurrencies
• Structure from motion (SfM) software

CUDA competes with other GPU computing stacks: Intel OneAPI and AMD ROCm.

Whereas Nvidia's CUDA is closed-source, Intel's OneAPI and AMD's ROCm are open source.

• Array programming
• BrookGPU – the Stanford University graphics group's compiler
• CUDA binary (cubin) – a type of fat binary
• Molecular modeling on GPUs
• Numerical Library Collection – by NEC for their vector processor
• OptiX – ray tracing API by NVIDIA
• Parallel computing
• rCUDA – an API for computing on remote computers
• Stream processing
• SYCL – an open standard from Khronos Group for programming a variety of platforms, including GPUs, with single-source modern C++, similar to higher-level CUDA Runtime API (single-source)
• Vulkan – low-level, high-performance 3D graphics and computing API

• Buck, Ian; Foley, Tim; Horn, Daniel; Sugerman, Jeremy; Fatahalian, Kayvon; Houston, Mike; Hanrahan, Pat (2004-08-01). "Brook for GPUs: stream computing on graphics hardware". ACM Transactions on Graphics. 23 (3): 777–786. doi:10.1145/1015706.1015800. ISSN 0730-0301.
• Nickolls, John; Buck, Ian; Garland, Michael; Skadron, Kevin (2008-03-01). "Scalable Parallel Programming with CUDA: Is CUDA the parallel programming model that application developers have been waiting for?". ACM Queue. 6 (2): 40–53. doi:10.1145/1365490.1365500. ISSN 1542-7730.

• Official website
• CUDA textbooks