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 13.4 samples update - v13.4-public

Browse Source 
Release 13.4 of the CUDA Samples supported by CUDA Toolkit 13.4.
See Changelog for more information.
This commit is contained in:
Dheemanth
2026-09-09 15:07:08 -07:00
committed by GitHub
parent b7c5481c55
commit 5443602d89
871 changed files with 6843 additions and 321849 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
cpp/3_CUDA_Features/cdpAdvancedQuicksort/.vscode/c_cpp_properties.json vendored
View File

@@ -1,18 +0,0 @@
{
"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/cdpAdvancedQuicksort/.vscode/extensions.json vendored
View File

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

43
cpp/3_CUDA_Features/cdpAdvancedQuicksort/CMakeLists.txt
View File

@@ -1,36 +1,28 @@
cmake_minimum_required(VERSION 3.20)
list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/Modules")
# CDP device-links the CUDA device runtime (cudadevrt), which the DriveOS Tegra
# toolkit ships only for the native Tegra archs; drop the desktop archs there.
get_filename_component(_ctk_root "${CMAKE_CUDA_COMPILER}" DIRECTORY)
get_filename_component(_ctk_root "${_ctk_root}" DIRECTORY)
if(EXISTS "${_ctk_root}/targets/aarch64-linux" OR EXISTS "${_ctk_root}/targets/aarch64-qnx")
set(SAMPLE_DISALLOW_ARCHS 75 80 86 89 90 100 103 107 120)
endif()
unset(_ctk_root)
include("${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/DetectCudaArch.cmake")
if(SAMPLE_SKIP_BUILD)
return()
endif()
project(cdpAdvancedQuicksort LANGUAGES C CXX CUDA)
include("${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/CudaSampleCommon.cmake")
find_package(CUDAToolkit REQUIRED)
set(CMAKE_POSITION_INDEPENDENT_CODE ON)
# The aarch64/sbsa_aarch64 CUDA toolkit are support on Tegra since 13.0, so need to check which version of the toolkit is installed
string(FIND "${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 "${CMAKE_CUDA_FLAGS} -G") # enable cuda-gdb (may significantly affect performance on some targets)
set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} -maxrregcount=64") # Limit register usage to 64 for the 'big_bitonicsort kernel
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 cdpAdvancedQuicksort
add_executable(cdpAdvancedQuicksort cdpAdvancedQuicksort.cu cdpBitonicSort.cu)
target_compile_options(cdpAdvancedQuicksort PRIVATE $<$<COMPILE_LANGUAGE:CUDA>:--extended-lambda>)
@@ -39,6 +31,5 @@ target_compile_features(cdpAdvancedQuicksort PRIVATE cxx_std_17 cuda_std_17)
set_target_properties(cdpAdvancedQuicksort PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
# Include installation configuration
include(${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/InstallSamples.cmake)
include(InstallSamples)
setup_samples_install()

4
cpp/3_CUDA_Features/cdpAdvancedQuicksort/cdpAdvancedQuicksort.cu
View File

@@ -137,7 +137,9 @@ template <typename T> static __device__ void ringbufFree(qsortRingbuf *ringbuf,
// and cover the instruction overhead.
//
////////////////////////////////////////////////////////////////////////////////
__global__ void qsort_warp(unsigned *indata,
// __launch_bounds__ caps registers to the launch block size so -G (device-debug) builds stay
// within the per-block register limit (big register usage under -G overflows it on Blackwell sm_110+).
__launch_bounds__(QSORT_BLOCKSIZE) __global__ void qsort_warp(unsigned *indata,
unsigned *outdata,
unsigned int offset,
unsigned int len,

6
cpp/3_CUDA_Features/cdpAdvancedQuicksort/cdpBitonicSort.cu
View File

@@ -272,14 +272,16 @@ static __device__ __forceinline__ void big_bitonicsort_kernel(unsigned *i
// KERNELS
////////////////////////////////////////////////////////////////////////////////
__global__ void bitonicsort(unsigned *indata, unsigned *outdata, unsigned int offset, unsigned int len)
// __launch_bounds__ caps registers to the launch block size so -G builds stay within the per-block register limit (Blackwell sm_110+).
__launch_bounds__(BITONICSORT_LEN) __global__ void bitonicsort(unsigned *indata, unsigned *outdata, unsigned int offset, unsigned int len)
{
// Handle to thread block group
cg::thread_block cta = cg::this_thread_block();
bitonicsort_kernel(indata, outdata, offset, len, cta);
}
__global__ void
// __launch_bounds__ caps registers to the launch block size so -G builds stay within the per-block register limit (Blackwell sm_110+).
__launch_bounds__(BITONICSORT_LEN) __global__ void
big_bitonicsort(unsigned *indata, unsigned *outdata, unsigned *backbuf, unsigned int offset, unsigned int len)
{
// Handle to thread block group