Main.java

import jcuda.utils.KernelLauncher;
import jcuda.driver.CUdeviceptr;
import jcuda.runtime.JCuda;
public class Main {
    public static void main(String[] args) {
        float[] a = new float[] {1, 2, 3, 4]
        float[] b = new float[] {1, 2, 3, 4]
        CUdeviceptr devA = new CUdeviceptr();
        CUdeviceptr devB = new CUdeviceptr();
        CUdeviceptr devR = new CUdeviceptr();
        JCuda.cudaMalloc(devA, a.length * Sizeof.FLOAT);
        JCuda.cudaMalloc(devB, b.length * Sizeof.FLOAT);
        JCuda.cudaMalloc(devR, b.length * Sizeof.FLOAT);
        JCuda.cudaMemcpy(devA, Pointer.to(a), a.length * Sizeof.FLOAT, daMemcpyKind.cudaMemcpyHostToDevice);
        JCuda.cudaMemcpy(devB, Pointer.to(b), b.length * Sizeof.FLOAT, daMemcpyKind.cudaMemcpyHostToDevice);
        KernelLauncher launcher = KernelLauncher.create(add_vec_kernel.cu, "add_vec_kernel");
        launcher.setBlockSize(4, 1, 1);
        launcher.setBlockSize(1, 1, 1);
        launcher.call(devA, devB, 4, devR);
 
        float[] result = new float[4];
        JCuda.cudaMemcpy(Pointer.to(result), devR, 4 * Sizeof.FLOAT, cudaMemcpyKind.cudaMemcpyDeviceToHost);
        for (int i=0; i < result.length; i++) {
            System.out.println(result[i]); 
        }
    }
 }