nvhpc

Link to section 'Description' of 'nvhpc' Description

The NVIDIA HPC SDK C, C++, and Fortran compilers support GPU acceleration of HPC modeling and simulation applications with standard C++ and Fortran, OpenACC® directives, and CUDA®. GPU-accelerated math libraries maximize performance on common HPC algorithms, and optimized communications libraries enable standards-based multi-GPU and scalable systems programming.

Link to section 'Homepage' of 'nvhpc' Homepage

https://developer.nvidia.com/hpc-sdk

Link to section 'Versions' of 'nvhpc' Versions

Scholar: 22.11
Gilbreth: 22.7
Anvil: 21.7

Link to section 'Module' of 'nvhpc' Module

You can load the modules by:

module purge
module load nvhpc

Link to section 'Example' of 'nvhpc' Example

Below is the example to use nvcc to compile a simple hello-world cuda code.

Link to section 'Cuda code' of 'nvhpc' Cuda code

hello.cu

#include <stdio.h>


__global__ void cuda_hello(){
    printf("Hello World from GPU!\n");
}

int main() {
    cuda_hello<<<1,1>>>();
    cudaDeviceSynchronize();
    return 0;
}
</stdio.h>

Link to section 'Compile cuda code' of 'nvhpc' Compile cuda code

$ nvcc hello.cu -o hello

Link to section 'Run compiled code' of 'nvhpc' Run compiled code

$ ./hello
  Hello World from GPU!