for (int i = 0; i < h; i++) {
    for (int j = 0; j < w; j++) 
        m[i*w+j] = i * w + j;
}

for (int i = 0; i < h; i++) {
    DWORD *first = &m[i*w];
    DWORD *last = &m[(i+1)*w-1];
    while (first <= last) 
        *first++ = i;
}

for (int i = 0; i < h; i++) {
V>    for (int j = 0; j < w; j++) 
V>        m[i*w+j] = i * w + j;
V>}

for (int i = 0; i < h; i++) {
V>    DWORD *first = &m[i*w];
V>    DWORD *last = &m[(i+1)*w-1];
V>    while (first <= last) 
V>        *first++ = i;
V>}

#include <chrono>
#include <iostream>

int main(int argc, char const *argv[]) {
    std::chrono::steady_clock::time_point begin, end;
    double dt, size, gb=1024*1024*1024;
    
    enum { N=16384 };
    enum { m=2048 };
    typedef unsigned T;
    T* x=new T[N*N]; if (!x) return 1;
    size=N*N*sizeof(*x);
    begin = std::chrono::steady_clock::now();
    for(int i=0;i<N;i++) {
        for(int j=0;j<N;j+=m) {
            for(int k=0;k<m;k++) x[i*N+j+k]=0; // -30%
            for(int k=0;k<m;k++) x[i*N+j+k]=i;
        }
    }
    end = std::chrono::steady_clock::now();
    dt=std::chrono::duration<double>(end-begin).count();
    std::cout << "cold dt = " << std::chrono::duration_cast<std::chrono::milliseconds>(end-begin).count() << "[ms]";
    std::cout << " speed=" << size/dt/gb << " Gb/s\n";
    begin = std::chrono::steady_clock::now();
    for(int i=0;i<N;i++) {
        for(int j=0;j<N;j+=m) {
            for(int k=0;k<m;k++) x[i*N+j+k]=0; // +90% :)
            for(int k=0;k<m;k++) x[i*N+j+k]=i;
        }
    }
    end = std::chrono::steady_clock::now();
    dt=std::chrono::duration<double>(end-begin).count();
    std::cout << "warm dt = " << std::chrono::duration_cast<std::chrono::milliseconds> (end - begin).count() << "[ms]";
    std::cout << " speed=" << size/dt/gb << " Gb/s\n";
    delete[] x;
    return 0;
}

$ g++ a.cpp -O3 && ./a.out
cold dt = 0[ms] speed=8.27842e+06 Gb/s
warm dt = 0[ms] speed=3.72529e+07 Gb/s

$ g++ a.cpp -O2 && ./a.out
cold dt = 0[ms] speed=7.45058e+06 Gb/s
warm dt = 0[ms] speed=3.72529e+07 Gb/s

$ g++ a.cpp -O1 && ./a.out
cold dt = 88[ms] speed=8.39206 Gb/s
warm dt = 62[ms] speed=11.9538 Gb/s

$ g++ a.cpp -O0 && ./a.out
cold dt = 296[ms] speed=2.50938 Gb/s
warm dt = 273[ms] speed=2.72459 Gb/s