56fd68e7dd
CCCL (CUDA C++ Core Libraries) provides: - CUB: device/block/warp-level GPU primitives (reduce, scan, sort, topk) - Thrust: high-level parallel algorithms (transform_reduce, sort, scan) - libcudacxx: CUDA C++ standard library (atomics, barriers, memory) - cudax: experimental features (memory resources, allocators) - Tuning policies: per-SM hardware-specific algorithm parameters Competition optimization vectors mapped to CCCL: - Output TPS (83% weight): warp_reduce, block_reduce, device_topk - Input TPS (14% weight): device_scan, block_load, prefetch - Cache TPS (3% weight): prefix caching strategy patterns - Memory (0.9 util): pooled/cached/buddy allocators Source: https://github.com/NVIDIA/cccl (shallow clone, HEAD only) License: Apache-2.0
128 lines
4.0 KiB
Plaintext
128 lines
4.0 KiB
Plaintext
/* 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.
|
|
*/
|
|
|
|
/**
|
|
* Vector addition: C = A + B.
|
|
*
|
|
* This sample is a very basic sample that implements element by element
|
|
* vector addition. It is the same as the sample illustrating Chapter 2
|
|
* of the programming guide with some additions like error checking.
|
|
*/
|
|
|
|
#include <cstdio>
|
|
|
|
// For the CUDA runtime routines (prefixed with "cuda_")
|
|
#include <cuda/std/span>
|
|
|
|
#include <cuda/experimental/launch.cuh>
|
|
#include <cuda/experimental/stream.cuh>
|
|
|
|
#include <cuda_runtime.h>
|
|
|
|
#include "vector.cuh"
|
|
|
|
namespace cudax = cuda::experimental;
|
|
using cudax::in;
|
|
using cudax::out;
|
|
|
|
/**
|
|
* CUDA Kernel Device code
|
|
*
|
|
* Computes the vector addition of A and B into C. The 3 vectors have the same
|
|
* number of elements numElements.
|
|
*/
|
|
__global__ void vectorAdd(cudax::span<const float> A, cudax::span<const float> B, cudax::span<float> C)
|
|
{
|
|
int i = static_cast<int>(blockDim.x * blockIdx.x + threadIdx.x);
|
|
|
|
if (i < A.size())
|
|
{
|
|
C[i] = A[i] + B[i] + 0.0f;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Host main routine
|
|
*/
|
|
int main()
|
|
try
|
|
{
|
|
// A CUDA stream on which to execute the vector addition kernel
|
|
cudax::stream stream(cuda::devices[0]);
|
|
|
|
// Print the vector length to be used, and compute its size
|
|
int numElements = 50000;
|
|
printf("[Vector addition of %d elements]\n", numElements);
|
|
|
|
// Allocate the host vectors
|
|
cudax::vector<float> A(numElements); // input
|
|
cudax::vector<float> B(numElements); // input
|
|
cudax::vector<float> C(numElements); // output
|
|
|
|
// Initialize the host input vectors
|
|
for (int i = 0; i < numElements; ++i)
|
|
{
|
|
A[i] = static_cast<float>(rand()) / (float) RAND_MAX;
|
|
B[i] = static_cast<float>(rand()) / (float) RAND_MAX;
|
|
}
|
|
|
|
// Define the kernel launch parameters
|
|
constexpr int threadsPerBlock = 256;
|
|
auto config = cuda::distribute<threadsPerBlock>(numElements);
|
|
|
|
// Launch the vectorAdd kernel
|
|
printf("CUDA kernel launch with %zu blocks of %d threads\n", cuda::block.count(cuda::grid, config), threadsPerBlock);
|
|
cudax::launch(stream, config, vectorAdd, in(A), in(B), out(C));
|
|
|
|
printf("waiting for the stream to finish\n");
|
|
stream.sync();
|
|
|
|
printf("verifying the results\n");
|
|
// Verify that the result vector is correct
|
|
for (int i = 0; i < numElements; ++i)
|
|
{
|
|
if (fabs(A[i] + B[i] - C[i]) > 1e-5)
|
|
{
|
|
fprintf(stderr, "Result verification failed at element %d!\n", i);
|
|
exit(EXIT_FAILURE);
|
|
}
|
|
}
|
|
|
|
printf("Test PASSED\n");
|
|
|
|
printf("Done\n");
|
|
return 0;
|
|
}
|
|
catch (const std::exception& e)
|
|
{
|
|
printf("caught an exception: \"%s\"\n", e.what());
|
|
}
|
|
catch (...)
|
|
{
|
|
printf("caught an unknown exception\n");
|
|
}
|