Как писать на ассемблере в 2021 году

#include <stdio.h>
#include <stdlib.h>
#define _USE_MATH_DEFINES
#include <math.h>
#include <chrono>

void ht_cpp0(double* data, double* result, int n)
{
    double phi = 2.0 * M_PI / n;
    for (int i = 0; i < n; ++i)
    {
        double sum = 0.0;
        for (int j = 0; j < n; ++j)
        {
            double w = phi * i * j;
            sum += data[j] * (cos(w) + sin(w));
        }
        result[i] = sum / sqrt(n);
    }
}

void ht_cpp1(double* data, double* result, int n)
{
    double phi = 2.0 * M_PI / n;
    double cos_phi = cos(phi);
    double sin_phi = sin(phi);
    double sqrtn = 1 / sqrt(n);
    double sin_dw = 0.0;
    double cos_dw = 1.0;
    for (int i = 0; i < n; ++i)
    {
        double sin_w = 0.0;
        double cos_w = 1.0;
        double sum = 0.0;

        for (int j = 0; j < n; ++j)
        {
            sum += data[j] * (cos_w + sin_w);
            double new_cos_w = cos_w * cos_dw - sin_w * sin_dw;
            double new_sin_w = sin_w * cos_dw + cos_w * sin_dw;
            cos_w = new_cos_w;
            sin_w = new_sin_w;
        }

        result[i] = sum * sqrtn;
        double new_cos_dw = cos_dw * cos_phi - sin_dw * sin_phi;
        double new_sin_dw = sin_dw * cos_phi + cos_dw * sin_phi;
        cos_dw = new_cos_dw;
        sin_dw = new_sin_dw;
    }
}

double MeanSquaredError(double* data1, double* data2, int n)
{
    double sum = 0;
    for (int i = 0; i < n; i++)
    {
        double df = data1[i] - data2[i];
        sum += df * df;
    }
    return sqrt(sum);
}

int main()
{
    enum : int { data_length = 10000 };
    double data_src[data_length];
    for (int i = 0; i < data_length; i++)
        data_src[i] = i + 1;
    double data_in[data_length];
    double data_out[data_length];

    {
        auto t0 = std::chrono::high_resolution_clock::now();
        ht_cpp0(data_src, data_in, data_length);
        ht_cpp0(data_in, data_out, data_length);
        auto t1 = std::chrono::high_resolution_clock::now();
        std::chrono::duration<double> ms_double = t1 - t0;
        double err = MeanSquaredError(data_src, data_out, data_length);
        printf("cpp0 : %lf\terr : %1.10lf\n", ms_double, err);
    }

    {
        auto t0 = std::chrono::high_resolution_clock::now();
        ht_cpp1(data_src, data_in, data_length);
        ht_cpp1(data_in, data_out, data_length);
        auto t1 = std::chrono::high_resolution_clock::now();
        std::chrono::duration<double> ms_double = t1 - t0;
        double err = MeanSquaredError(data_src, data_out, data_length);
        printf("cpp1 : %lf\terr : %1.10lf\n", ms_double, err);
    }

    system("pause");
}