NASP/cuda-samples
84
0
Fork 0
mirror of https://github.com/NVIDIA/cuda-samples.git synced 2026-10-13 00:08:26 +08:00
Code Issues Packages Projects Releases Wiki Activity

CUDA 13.2 samples update (#432)

Browse Source 
- Added Python samples for CUDA Python 1.0 release
- Renamed top-level `Samples` directory to `cpp` to accommodate Python samples.
This commit is contained in:
Dheemanth
2026-05-13 15:13:18 -07:00
committed by GitHub
parent a4526d5229
commit aeab82ff30
2072 changed files with 15077 additions and 1970 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
cpp/3_CUDA_Features/cdpBezierTessellation/.vscode/c_cpp_properties.json vendored Normal file
View File

@@ -0,0 +1,18 @@
{
"configurations": [
{
"name": "Linux",
"includePath": [
"${workspaceFolder}/**",
"${workspaceFolder}/../../../Common"
],
"defines": [],
"compilerPath": "/usr/local/cuda/bin/nvcc",
"cStandard": "gnu17",
"cppStandard": "gnu++14",
"intelliSenseMode": "linux-gcc-x64",
"configurationProvider": "ms-vscode.makefile-tools"
}
],
"version": 4
}

7
cpp/3_CUDA_Features/cdpBezierTessellation/.vscode/extensions.json vendored Normal file
View File

@@ -0,0 +1,7 @@
{
"recommendations": [
"nvidia.nsight-vscode-edition",
"ms-vscode.cpptools",
"ms-vscode.makefile-tools"
]
}

205
cpp/3_CUDA_Features/cdpBezierTessellation/BezierLineCDP.cu Normal file
View File

@@ -0,0 +1,205 @@
/* Copyright (c) 2022, NVIDIA CORPORATION. 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 NVIDIA CORPORATION nor the names of its
* 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 ``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.
*/
#include <cuda_runtime_api.h>
#include <helper_cuda.h>
#include <stdio.h>
#include <string.h>
__forceinline__ __device__ float2 operator+(float2 a, float2 b)
{
float2 c;
c.x = a.x + b.x;
c.y = a.y + b.y;
return c;
}
__forceinline__ __device__ float2 operator-(float2 a, float2 b)
{
float2 c;
c.x = a.x - b.x;
c.y = a.y - b.y;
return c;
}
__forceinline__ __device__ float2 operator*(float a, float2 b)
{
float2 c;
c.x = a * b.x;
c.y = a * b.y;
return c;
}
__forceinline__ __device__ float length(float2 a) { return sqrtf(a.x * a.x + a.y * a.y); }
#define MAX_TESSELLATION 32
struct BezierLine
{
float2 CP[3];
float2 *vertexPos;
int nVertices;
};
__global__ void computeBezierLinePositions(int lidx, BezierLine *bLines, int nTessPoints)
{
int idx = threadIdx.x + blockDim.x * blockIdx.x;
if (idx < nTessPoints) {
float u = (float)idx / (float)(nTessPoints - 1);
float omu = 1.0f - u;
float B3u[3];
B3u[0] = omu * omu;
B3u[1] = 2.0f * u * omu;
B3u[2] = u * u;
float2 position = {0, 0};
for (int i = 0; i < 3; i++) {
position = position + B3u[i] * bLines[lidx].CP[i];
}
bLines[lidx].vertexPos[idx] = position;
}
}
__global__ void computeBezierLinesCDP(BezierLine *bLines, int nLines)
{
int lidx = threadIdx.x + blockDim.x * blockIdx.x;
if (lidx < nLines) {
float curvature = length(bLines[lidx].CP[1] - 0.5f * (bLines[lidx].CP[0] + bLines[lidx].CP[2]))
/ length(bLines[lidx].CP[2] - bLines[lidx].CP[0]);
int nTessPoints = min(max((int)(curvature * 16.0f), 4), MAX_TESSELLATION);
if (bLines[lidx].vertexPos == NULL) {
bLines[lidx].nVertices = nTessPoints;
cudaMalloc((void **)&bLines[lidx].vertexPos, nTessPoints * sizeof(float2));
}
computeBezierLinePositions<<<ceilf((float)bLines[lidx].nVertices / 32.0f), 32>>>(
lidx, bLines, bLines[lidx].nVertices);
}
}
__global__ void freeVertexMem(BezierLine *bLines, int nLines)
{
int lidx = threadIdx.x + blockDim.x * blockIdx.x;
if (lidx < nLines)
cudaFree(bLines[lidx].vertexPos);
}
unsigned int checkCapableSM35Device(int argc, char **argv)
{
// Get device properties
cudaDeviceProp properties;
int device_count = 0, device = -1;
if (checkCmdLineFlag(argc, (const char **)argv, "device")) {
device = getCmdLineArgumentInt(argc, (const char **)argv, "device");
cudaDeviceProp properties;
checkCudaErrors(cudaGetDeviceProperties(&properties, device));
if (properties.major > 3 || (properties.major == 3 && properties.minor >= 5)) {
printf("Running on GPU %d (%s)\n", device, properties.name);
}
else {
printf("cdpBezierTessellation requires GPU devices with compute SM 3.5 or "
"higher.");
printf("Current GPU device has compute SM %d.%d. Exiting...\n", properties.major, properties.minor);
return EXIT_FAILURE;
}
}
else {
checkCudaErrors(cudaGetDeviceCount(&device_count));
for (int i = 0; i < device_count; ++i) {
checkCudaErrors(cudaGetDeviceProperties(&properties, i));
if (properties.major > 3 || (properties.major == 3 && properties.minor >= 5)) {
device = i;
printf("Running on GPU %d (%s)\n", i, properties.name);
break;
}
printf("GPU %d %s does not support CUDA Dynamic Parallelism\n", i, properties.name);
}
}
if (device == -1) {
fprintf(stderr,
"cdpBezierTessellation requires GPU devices with compute SM 3.5 or "
"higher. Exiting...\n");
return EXIT_WAIVED;
}
return EXIT_SUCCESS;
}
#define N_LINES 256
#define BLOCK_DIM 64
int main(int argc, char **argv)
{
BezierLine *bLines_h = new BezierLine[N_LINES];
float2 last = {0, 0};
for (int i = 0; i < N_LINES; i++) {
bLines_h[i].CP[0] = last;
for (int j = 1; j < 3; j++) {
bLines_h[i].CP[j].x = (float)rand() / (float)RAND_MAX;
bLines_h[i].CP[j].y = (float)rand() / (float)RAND_MAX;
}
last = bLines_h[i].CP[2];
bLines_h[i].vertexPos = NULL;
bLines_h[i].nVertices = 0;
}
unsigned int sm35Ret = checkCapableSM35Device(argc, argv);
if (sm35Ret != EXIT_SUCCESS) {
exit(sm35Ret);
}
BezierLine *bLines_d;
checkCudaErrors(cudaMalloc((void **)&bLines_d, N_LINES * sizeof(BezierLine)));
checkCudaErrors(cudaMemcpy(bLines_d, bLines_h, N_LINES * sizeof(BezierLine), cudaMemcpyHostToDevice));
printf("Computing Bezier Lines (CUDA Dynamic Parallelism Version) ... ");
computeBezierLinesCDP<<<(unsigned int)ceil((float)N_LINES / (float)BLOCK_DIM), BLOCK_DIM>>>(bLines_d, N_LINES);
printf("Done!\n");
// Do something to draw the lines here
freeVertexMem<<<(unsigned int)ceil((float)N_LINES / (float)BLOCK_DIM), BLOCK_DIM>>>(bLines_d, N_LINES);
checkCudaErrors(cudaFree(bLines_d));
delete[] bLines_h;
exit(EXIT_SUCCESS);
}

43
cpp/3_CUDA_Features/cdpBezierTessellation/CMakeLists.txt Normal file
View File

@@ -0,0 +1,43 @@
cmake_minimum_required(VERSION 3.20)
list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/Modules")
project(cdpBezierTessellation LANGUAGES C CXX CUDA)
find_package(CUDAToolkit REQUIRED)
set(CMAKE_POSITION_INDEPENDENT_CODE ON)
# The aarch64/sbsa_aarch64 CUDA toolkit are support on Tegra since 13.0, so need to check which version of the toolkit is installed
string(FIND "${CUDAToolkit_INCLUDE_DIRS}" "aarch64-linux" _aarch64_linux_ctk)
string(FIND "${CUDAToolkit_INCLUDE_DIRS}" "aarch64-qnx" _aarch64_qnx_ctk)
if(CMAKE_SYSTEM_PROCESSOR STREQUAL "aarch64" AND (NOT _aarch64_linux_ctk EQUAL -1 OR NOT _aarch64_qnx_ctk EQUAL -1))
set(CMAKE_CUDA_ARCHITECTURES 87 110)
else()
set(CMAKE_CUDA_ARCHITECTURES 75 80 86 89 90 100 110 120)
endif()
set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -Wno-deprecated-gpu-targets")
if(ENABLE_CUDA_DEBUG)
set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -G") # enable cuda-gdb (may significantly affect performance on some targets)
else()
set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -lineinfo") # add line information to all builds for debug tools (exclusive to -G option)
endif()
# Include directories and libraries
include_directories(../../../Common)
# Source file
# Add target for cdpBezierTessellation
add_executable(cdpBezierTessellation BezierLineCDP.cu)
target_compile_options(cdpBezierTessellation PRIVATE $<$<COMPILE_LANGUAGE:CUDA>:--extended-lambda>)
target_compile_features(cdpBezierTessellation PRIVATE cxx_std_17 cuda_std_17)
set_target_properties(cdpBezierTessellation PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
# Include installation configuration
include(${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/InstallSamples.cmake)
setup_samples_install()

34
cpp/3_CUDA_Features/cdpBezierTessellation/README.md Normal file
View File

@@ -0,0 +1,34 @@
# cdpBezierTessellation - Bezier Line Tessellation (CUDA Dynamic Parallelism)
## Description
This sample demonstrates bezier tessellation of lines implemented using CUDA Dynamic Parallelism. This sample requires devices with compute capability 3.5 or higher.
## Key Concepts
CUDA Dynamic Parallelism
## Supported SM Architectures
## Supported OSes
Linux, Windows
## Supported CPU Architecture
x86_64, armv7l
## CUDA APIs involved
### [CUDA Runtime API](http://docs.nvidia.com/cuda/cuda-runtime-api/index.html)
cudaMemcpy, cudaFree, cudaGetDeviceCount, cudaMalloc, cudaGetDeviceProperties
## Dependencies needed to build/run
[CDP](../../../README.md#cuda-dynamic-parallellism)
## Prerequisites
Download and install the [CUDA Toolkit](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
Make sure the dependencies mentioned in [Dependencies]() section above are installed.
## References (for more details)