Data Locality: Matmul

matmul_a.cpp

void A(int n, int** a, int** b, int** c){
	parallel_for(0, n, [&](int k)
	{
		for (int i = 0; i<n; i++) {
			int r = a[i][k];
			for (int j = 0; j<n; j++)
				c[i][j] += r * b[k][j];
		}
	});
};

matmul_b.cpp

void Parallel_KJI(int n, int** a, int** b, int** c){
	parallel_for(0, n, [&](int k)
	{
		for (int j = 0; j<n; j++) {
			int r = b[k][j];
			for (int i = 0; i<n; i++)
				c[i][j] += a[i][k] * r;
		}
	});
};