CUDA Standard Algorithms » Single Task

Include the Header

You need to include the header file, taskflow/cuda/algorithm/for_each.hpp, for creating a single-threaded task.

#include <taskflow/cuda/algorithm/for_each.hpp>

Run a Task with a Single Thread

You can launch a kernel with only one GPU thread running it, which is handy when you want to set up a single or a few variables that do not need multiple threads. The following example creates a single-task kernel that sets a device variable to 1.

tf::cudaStream stream;
tf::cudaDefaultExecutionPolicy policy(stream);

// launch the single-task kernel asynchronously through the policy
tf::cuda_single_task(policy, [gpu_variable] __device__ () {
  *gpu_Variable = 1;
});

// wait for the kernel completes
stream.synchronize();

Since the callable runs on GPU, it must be declared with a __device__ specifier.