const int nStreams = 4;
cudaStream_t streams[nStreams];
for (int i = 0; i < nStreams; ++i) cudaStreamCreate(&streams[i]);

int chunkSize = n / nStreams;
size_t chunkBytes = chunkSize * sizeof(float);

for (int i = 0; i < nStreams; ++i) {
    int offset = i * chunkSize;
    cudaMemcpyAsync(d_data + offset, h_data + offset, chunkBytes,
                     cudaMemcpyHostToDevice, streams[i]);
    myKernel<<<chunkSize / 256, 256, 0, streams[i]>>>(d_data + offset, chunkSize);
    cudaMemcpyAsync(h_result + offset, d_result + offset, chunkBytes,
                     cudaMemcpyDeviceToHost, streams[i]);
}

for (int i = 0; i < nStreams; ++i) cudaStreamSynchronize(streams[i]);
for (int i = 0; i < nStreams; ++i) cudaStreamDestroy(streams[i]);
