CUDA 13.2 samples update (#432)

- Added Python samples for CUDA Python 1.0 release
- Renamed top-level `Samples` directory to `cpp` to accommodate Python samples.

cpp/2_Concepts_and_Techniques/radixSortThrust/.vscode/c_cpp_properties.json

@@ -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
+}

cpp/2_Concepts_and_Techniques/radixSortThrust/.vscode/extensions.json

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

cpp/2_Concepts_and_Techniques/radixSortThrust/CMakeLists.txt

@@ -0,0 +1,34 @@
+cmake_minimum_required(VERSION 3.20)
+
+list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/Modules")
+
+project(radixSortThrust LANGUAGES C CXX CUDA)
+
+find_package(CUDAToolkit REQUIRED)
+
+set(CMAKE_POSITION_INDEPENDENT_CODE ON)
+
+set(CMAKE_CUDA_ARCHITECTURES 75 80 86 87 89 90 100 110 120)
+set(CMAKE_CUDA_FLAGS "${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 radixSortThrust
+add_executable(radixSortThrust radixSortThrust.cu)
+
+target_compile_options(radixSortThrust PRIVATE $<$<COMPILE_LANGUAGE:CUDA>:--extended-lambda>)
+
+target_compile_features(radixSortThrust PRIVATE cxx_std_17 cuda_std_17)
+
+set_target_properties(radixSortThrust PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
+
+# Include installation configuration
+include(${CMAKE_CURRENT_SOURCE_DIR}/../../../cmake/InstallSamples.cmake)
+setup_samples_install()

cpp/2_Concepts_and_Techniques/radixSortThrust/README.md

@@ -0,0 +1,34 @@
+# radixSortThrust - CUDA Radix Sort (Thrust Library)
+
+## Description
+
+This sample demonstrates a very fast and efficient parallel radix sort uses Thrust library. The included RadixSort class can sort either key-value pairs (with float or unsigned integer keys) or keys only.
+
+## Key Concepts
+
+Data-Parallel Algorithms, Performance Strategies
+
+## Supported SM Architectures
+
+[SM 5.0 ](https://developer.nvidia.com/cuda-gpus)  [SM 5.2 ](https://developer.nvidia.com/cuda-gpus)  [SM 5.3 ](https://developer.nvidia.com/cuda-gpus)  [SM 6.0 ](https://developer.nvidia.com/cuda-gpus)  [SM 6.1 ](https://developer.nvidia.com/cuda-gpus)  [SM 7.0 ](https://developer.nvidia.com/cuda-gpus)  [SM 7.2 ](https://developer.nvidia.com/cuda-gpus)  [SM 7.5 ](https://developer.nvidia.com/cuda-gpus)  [SM 8.0 ](https://developer.nvidia.com/cuda-gpus)  [SM 8.6 ](https://developer.nvidia.com/cuda-gpus)  [SM 8.7 ](https://developer.nvidia.com/cuda-gpus)  [SM 8.9 ](https://developer.nvidia.com/cuda-gpus)  [SM 9.0 ](https://developer.nvidia.com/cuda-gpus)
+
+## 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)
+cudaEventSynchronize, cudaEventRecord, cudaGetDevice, cudaEventDestroy, cudaEventElapsedTime, cudaGetDeviceProperties, cudaEventCreate
+
+## Prerequisites
+
+Download and install the [CUDA Toolkit](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
+
+## References (for more details)
+
+[whitepaper](./doc/readme.txt)

cpp/2_Concepts_and_Techniques/radixSortThrust/doc/readme.txt

@@ -0,0 +1,40 @@
+NVIDIA CUDA Sample "radixSortThrust"
+----------------------------------
+
+--------
+OVERVIEW
+--------
+
+This sample demonstrates a very fast and efficient parallel radix sort implemented in C for CUDA.  The included RadixSort class can sort either key-value pairs (with float or unsigned integer keys) or keys only. It can also sort unsigned integer keys based on a varying number of least-significant bits ranging from 4 to 32 in multiples of 4.
+
+This radix sort code and the underlying algorithm is discussed in detail in the paper "Designing Efficient Sorting Algorithms for Manycore GPUs".  A PDF version of this paper is available at http://mgarland.org/files/papers/gpusort-ipdps09.pdf
+
+-----
+USAGE
+-----
+
+To run a sort with default options (Sort 1M unsigned integer key-value pairs), just invoke the executable ("radixSort.exe" on Windows, "radixSort" otherwise).
+
+The following command line options are available:
+
+ -n=<N>        : number of elements to sort
+ -keysonly     : sort only an array of keys (the default is to sort key-value pairs)
+ -float        : use 32-bit float keys
+ -keybits=<B>  : Use only the B least-significant bits of the keys for the sort
+               : B must be a multiple of 4.  This option does not apply to float keys
+ -quiet        : Output only the number of elements and the time to sort
+ -help         : Output a help message
+
+The RadixSort class can also be used within your application by building the radixsort.cu file into your application or library, and including the radixsort.h header file.
+
+--------
+CITATION
+--------
+
+Satish, N., Harris, M., and Garland, M. "Designing Efficient Sorting
+Algorithms for Manycore GPUs". In Proceedings of IEEE International
+Parallel & Distributed Processing Symposium 2009 (IPDPS 2009).
+
+PDF:
+
+http://mgarland.org/files/papers/gpusort-ipdps09.pdf

cpp/2_Concepts_and_Techniques/radixSortThrust/radixSortThrust.cu

@@ -0,0 +1,217 @@
+/* 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 <algorithm>
+#include <helper_cuda.h>
+#include <limits.h>
+#include <thrust/copy.h>
+#include <thrust/detail/type_traits.h>
+#include <thrust/device_vector.h>
+#include <thrust/generate.h>
+#include <thrust/host_vector.h>
+#include <thrust/random.h>
+#include <thrust/sequence.h>
+#include <thrust/sort.h>
+#include <time.h>
+
+template <typename T, bool floatKeys> bool testSort(int argc, char **argv)
+{
+    int cmdVal;
+    int keybits = 32;
+
+    unsigned int numElements = 1048576;
+    bool         keysOnly    = checkCmdLineFlag(argc, (const char **)argv, "keysonly");
+    bool         quiet       = checkCmdLineFlag(argc, (const char **)argv, "quiet");
+
+    if (checkCmdLineFlag(argc, (const char **)argv, "n")) {
+        cmdVal      = getCmdLineArgumentInt(argc, (const char **)argv, "n");
+        numElements = cmdVal;
+
+        if (cmdVal < 0) {
+            printf("Error: elements must be > 0, elements=%d is invalid\n", cmdVal);
+            exit(EXIT_SUCCESS);
+        }
+    }
+
+    if (checkCmdLineFlag(argc, (const char **)argv, "keybits")) {
+        cmdVal  = getCmdLineArgumentInt(argc, (const char **)argv, "keybits");
+        keybits = cmdVal;
+
+        if (keybits <= 0) {
+            printf("Error: keybits must be > 0, keybits=%d is invalid\n", keybits);
+            exit(EXIT_SUCCESS);
+        }
+    }
+
+    unsigned int numIterations = (numElements >= 16777216) ? 10 : 100;
+
+    if (checkCmdLineFlag(argc, (const char **)argv, "iterations")) {
+        cmdVal        = getCmdLineArgumentInt(argc, (const char **)argv, "iterations");
+        numIterations = cmdVal;
+    }
+
+    if (checkCmdLineFlag(argc, (const char **)argv, "help")) {
+        printf("Command line:\nradixSortThrust [-option]\n");
+        printf("Valid options:\n");
+        printf("-n=<N>        : number of elements to sort\n");
+        printf("-keybits=bits : keybits must be > 0\n");
+        printf("-keysonly     : only sort an array of keys (default sorts key-value "
+               "pairs)\n");
+        printf("-float        : use 32-bit float keys (default is 32-bit unsigned "
+               "int)\n");
+        printf("-quiet        : Output only the number of elements and the time to "
+               "sort\n");
+        printf("-help         : Output a help message\n");
+        exit(EXIT_SUCCESS);
+    }
+
+    if (!quiet)
+        printf("\nSorting %d %d-bit %s keys %s\n\n",
+               numElements,
+               keybits,
+               floatKeys ? "float" : "unsigned int",
+               keysOnly ? "(only)" : "and values");
+
+    int deviceID = -1;
+
+    if (cudaSuccess == cudaGetDevice(&deviceID)) {
+        cudaDeviceProp devprop;
+        cudaGetDeviceProperties(&devprop, deviceID);
+        unsigned int totalMem = (keysOnly ? 2 : 4) * numElements * sizeof(T);
+
+        if (devprop.totalGlobalMem < totalMem) {
+            printf("Error: insufficient amount of memory to sort %d elements.\n", numElements);
+            printf("%d bytes needed, %d bytes available\n", (int)totalMem, (int)devprop.totalGlobalMem);
+            exit(EXIT_SUCCESS);
+        }
+    }
+
+    thrust::host_vector<T>            h_keys(numElements);
+    thrust::host_vector<T>            h_keysSorted(numElements);
+    thrust::host_vector<unsigned int> h_values;
+
+    if (!keysOnly)
+        h_values = thrust::host_vector<unsigned int>(numElements);
+
+    // Fill up with some random data
+    thrust::default_random_engine rng(clock());
+
+    if (floatKeys) {
+        thrust::uniform_real_distribution<float> u01(0, 1);
+
+        for (int i = 0; i < (int)numElements; i++)
+            h_keys[i] = u01(rng);
+    }
+    else {
+        thrust::uniform_int_distribution<unsigned int> u(0, UINT_MAX);
+
+        for (int i = 0; i < (int)numElements; i++)
+            h_keys[i] = u(rng);
+    }
+
+    if (!keysOnly)
+        thrust::sequence(h_values.begin(), h_values.end());
+
+    // Copy data onto the GPU
+    thrust::device_vector<T>            d_keys;
+    thrust::device_vector<unsigned int> d_values;
+
+    // run multiple iterations to compute an average sort time
+    cudaEvent_t start_event, stop_event;
+    checkCudaErrors(cudaEventCreate(&start_event));
+    checkCudaErrors(cudaEventCreate(&stop_event));
+
+    float totalTime = 0;
+
+    for (unsigned int i = 0; i < numIterations; i++) {
+        // reset data before sort
+        d_keys = h_keys;
+
+        if (!keysOnly)
+            d_values = h_values;
+
+        checkCudaErrors(cudaEventRecord(start_event, 0));
+
+        if (keysOnly)
+            thrust::sort(d_keys.begin(), d_keys.end());
+        else
+            thrust::sort_by_key(d_keys.begin(), d_keys.end(), d_values.begin());
+
+        checkCudaErrors(cudaEventRecord(stop_event, 0));
+        checkCudaErrors(cudaEventSynchronize(stop_event));
+
+        float time = 0;
+        checkCudaErrors(cudaEventElapsedTime(&time, start_event, stop_event));
+        totalTime += time;
+    }
+
+    totalTime /= (1.0e3f * numIterations);
+    printf("radixSortThrust, Throughput = %.4f MElements/s, Time = %.5f s, Size = "
+           "%u elements\n",
+           1.0e-6f * numElements / totalTime,
+           totalTime,
+           numElements);
+
+    getLastCudaError("after radixsort");
+
+    // Get results back to host for correctness checking
+    thrust::copy(d_keys.begin(), d_keys.end(), h_keysSorted.begin());
+
+    if (!keysOnly)
+        thrust::copy(d_values.begin(), d_values.end(), h_values.begin());
+
+    getLastCudaError("copying results to host memory");
+
+    // Check results
+    bool bTestResult = thrust::is_sorted(h_keysSorted.begin(), h_keysSorted.end());
+
+    checkCudaErrors(cudaEventDestroy(start_event));
+    checkCudaErrors(cudaEventDestroy(stop_event));
+
+    if (!bTestResult && !quiet) {
+        return false;
+    }
+
+    return bTestResult;
+}
+
+int main(int argc, char **argv)
+{
+    // Start logs
+    printf("%s Starting...\n\n", argv[0]);
+
+    findCudaDevice(argc, (const char **)argv);
+
+    bool bTestResult = false;
+
+    if (checkCmdLineFlag(argc, (const char **)argv, "float"))
+        bTestResult = testSort<float, true>(argc, argv);
+    else
+        bTestResult = testSort<unsigned int, false>(argc, argv);
+
+    printf(bTestResult ? "Test passed\n" : "Test failed!\n");
+}