Kernel Creation

Algorithm structures and patterns that create CUDA® GPU kernels

CUDA kernels are functions that are executed on the GPU device. These kernels are executed by many GPU threads in parallel. By efficiently mapping compute intensive portions of your algorithm to kernels, you can take advantage of the performance improvements provided by GPU computing technology. You can trigger GPU Coder™ to create CUDA kernels for specific algorithm structures and patterns in your MATLAB® code.

Apps

GPU CoderGenerate GPU code from MATLAB code
Check GPU InstallVerify and set up the GPU code generation environment

Functions

halfConstruct half-precision numeric object
coder.checkGpuInstallVerify GPU code generation environment
coder.gpuConfigConfiguration parameters for CUDA code generation from MATLAB code by using GPU Coder
codegenGenerate C/C++ code from MATLAB code
gpucoderOpen GPU Coder app
coder.gpu.kernelPragma that maps for-loops to GPU kernels
coder.gpu.kernelfunPragma that maps function to GPU kernels
coder.gpu.nokernelPragma to disable kernel creation for loops
coder.gpu.constantMemoryPragma that maps a variable to the constant memory on GPU
gpucoder.stencilKernelCreate CUDA code for stencil functions
gpucoder.matrixMatrixKernelOptimized GPU implementation of functions containing matrix-matrix operations
gpucoder.batchedMatrixMultiplyOptimized GPU implementation of batched matrix multiply operation
gpucoder.stridedMatrixMultiplyOptimized GPU implementation of strided and batched matrix multiply operation
gpucoder.batchedMatrixMultiplyAddOptimized GPU implementation of batched matrix multiply with add operation
gpucoder.stridedMatrixMultiplyAddOptimized GPU implementation of strided, batched matrix multiply with add operation
gpucoder.sortOptimized GPU implementation of the MATLAB sort function
coder.gpu.iterationsPragma that provides information to the code generator for making parallelization decisions on variable bound loops
gpucoder.transposeOptimized GPU implementation of the MATLAB transpose function
gpucoder.reduceOptimized GPU implementation for reduction operations
coder.cevalCall external C/C++ function

Objects

coder.gpuConfigConfiguration parameters for CUDA code generation from MATLAB code by using GPU Coder
coder.CodeConfigConfiguration parameters for C/C++ code generation from MATLAB code
coder.EmbeddedCodeConfigConfiguration parameters for C/C++ code generation from MATLAB code with Embedded Coder
coder.gpuEnvConfigCreate configuration object containing the parameters passed to coder.checkGpuInstall for performing GPU code generation environment checks

Topics

Kernels from Element-Wise Loops

Create kernels from MATLAB functions containing scalarized, element-wise math operations.

Kernels from Scatter-Gather Type Operations

Create kernels from MATLAB functions containing reduction operations.

Kernels from Library Calls

Target GPU optimized math libraries such as cuBLAS, cuSOLVER, cuFFT, and Thrust.

Support for GPU Arrays

Generate CUDA code that uses GPU arrays.

Legacy Code Integration

Integrate custom GPU code with MATLAB code intended for code generation.

Design Patterns

Create kernels for MATLAB functions containing computational design patterns.

GPU Memory Allocation and Minimization

Memory allocation options and optimizations for GPU Coder.

Featured Examples

Simulate Diffraction Patterns Using CUDA FFT Libraries

Simulate Diffraction Patterns Using CUDA FFT Libraries

Use GPU Coder™ to leverage the CUDA® Fast Fourier Transform library (cuFFT) to compute two-dimensional FFT on a NVIDIA® GPU. The two-dimensional Fourier transform is used in optics to calculate far-field diffraction patterns. When a monochromatic light source passes through a small aperture, such as in Young's double-slit experiment, you can observe these diffraction patterns. This example also shows you how to use GPU pointers as inputs to an entry-point function when generating CUDA MEX, source code, static libraries, dynamic libraries, and executables. By using this functionality, the performance of the generated code is improved by minimizing the number of cudaMemcpy calls in the generated code.

Open Script
QR Decomposition on NVIDIA GPU Using cuSOLVER Libraries

QR Decomposition on NVIDIA GPU Using cuSOLVER Libraries

Create a standalone CUDA® executable that leverages the CUDA Solver library (cuSOLVER). The example uses a curve fitting application that mimics automatic lane tracking on a road to illustrate: Fitting an arbitrary-order polynomial to noisy data by using matrix QR factorization.

Open Script
Stencil Processing on GPU

Stencil Processing on GPU

Generate CUDA® kernels for stencil type operations by implementing "Game of Life" by John H. Conway.

Open Script
Benchmark A\b by Using GPU Coder

Benchmark A\b by Using GPU Coder

Benchmark solving a linear system by generating GPU code. Use matrix left division, also known as mldivide or the backslash operator (\), to solve the system of linear equations $A*x = b$ for $x$ (that is, compute $x = A \backslash b$).

Open Script
Fog Rectification

Fog Rectification

The use of image processing functions for GPU code generation. The example takes a foggy image as input and produces a defogged image. This example is a typical implementation of fog rectification algorithm. The example uses conv2, rgb2gray, and imhist functions.

Open Script
Stereo Disparity

Stereo Disparity

Generate a MEX function from a MATLAB® function that computes the stereo disparity of two images.

Open Script
Feature Extraction Using SURF

Feature Extraction Using SURF

Object Recognition using Speeded-Up Robust Features (SURF) is composed of three steps: feature extraction, feature description, and feature matching. This example performs feature extraction, which is the first step of the SURF algorithm. The algorithm used here is based on the OpenSURF library implementation. This example shows how you can use GPU Coder™ to solve this compute intensive problem through CUDA® code generation.

Open Script
Feature Matching

Feature Matching

Generate CUDA® MEX from MATLAB® code and perform feature matching between two images. This example uses the matchFeatures function from the Image Processing Toolbox™ to match the feature descriptors between two images that are rotated and scaled with respect to each other. The feature descriptors of the two images are detected and extracted by using the Speeded-Up Robust Features (SURF) algorithm.

Open Script
Lane Detection on the GPU by Using the houghlines Function

Lane Detection on the GPU by Using the houghlines Function

Generate CUDA® MEX for a MATLAB® function that can detect and output lane marker boundaries on an image. The example takes an RGB image as input and uses the rgb2gray, ordfilt2, hough, houghpeaks, and houghlines functions that are part of Image Processing Toolbox™ to produce the lane-detected output image.

Open Script
Edge Detection with Sobel Method in Half-Precision

Edge Detection with Sobel Method in Half-Precision

Demonstrates edge detection in an image with a MEX function generated from a MATLAB® function. The edge detection algorithm is implemented with half-precision data type.

Open Script
GPU Coder Documentation
Support