Parrallel-Programming---Hw4---OpenMP/x86omp_SIMD.cpp at master · EmryZhang/Parrallel-Programming---Hw4---OpenMP · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
#include<iostream>
#include<time.h>
#include<pmmintrin.h>
#include<xmmintrin.h>
#include<immintrin.h>

#define N 1024
#define count 1
using namespace std;


float A[N][N];

void ori_LU() {
    for (int k = 0; k < N; k++){
        float tmp =A[k][k];
        for (int j = k; j < N; j++){
            A[k][j] = A[k][j] / tmp;
        }
        for (int i = k + 1; i < N; i++){
            float tmp2 = A[i][k];
            for (int j = k + 1; j < N; j++){
                A[i][j] = A[i][j] - tmp2 * A[k][j];
            }
            A[i][k] = 0;
        }
    }
}


void SSE_LU(){
    __m128 t1, t2, t3, t4;
    for (int k = 0; k < N; k++){
        float tmp[4] = { A[k][k], A[k][k], A[k][k], A[k][k] };
        t1 = _mm_loadu_ps(tmp);
        for (int j = N - 4; j >=k; j -= 4) {
            t2 = _mm_loadu_ps(A[k] + j);
            t3 = _mm_div_ps(t2, t1);
            _mm_storeu_ps(A[k] + j, t3);
        }
        if (k % 4 != (N % 4)){
            for (int j = k; j % 4 != ( N% 4); j++){
                A[k][j] = A[k][j] / tmp[0];
            }
        }
        for (int j = (N % 4) - 1; j>= 0; j--){
            A[k][j] = A[k][j] / tmp[0];
        }
        for (int i = k + 1; i < N; i++){
            float tmp[4] = { A[i][k], A[i][k], A[i][k], A[i][k] };
            t1 = _mm_loadu_ps(tmp);
            for (int j = N - 4; j >k;j -= 4){
                t2 = _mm_loadu_ps(A[i] + j);
                t3 = _mm_loadu_ps(A[k] + j);
                t4 = _mm_sub_ps(t2,_mm_mul_ps(t1, t3));
                _mm_storeu_ps(A[i] + j, t4);
            }
            for (int j = k + 1; j % 4 !=(N % 4); j++){
                A[i][j] = A[i][j] - A[i][k] * A[k][j];
            }
            A[i][k] = 0;
        }
    }
}

void AVX_LU(){
    __m256 t1, t2, t3, t4;
    for (int k = 0; k < N; k++){
        float tmp[8];
        for(int i=0;i<8;i++)tmp[i]=A[k][k];
        t1 = _mm256_loadu_ps(tmp);
        for (int j = N - 8; j >=k; j -= 8) {
            t2 = _mm256_loadu_ps(A[k] + j);
            t3 = _mm256_div_ps(t2, t1);
            _mm256_storeu_ps(A[k] + j, t3);
        }
        if (k % 8 != (N % 8)){
            for (int j = k; j % 8 != ( N% 8); j++){
                A[k][j] = A[k][j] / tmp[0];
            }
        }
        for (int j = (N % 8) - 1; j>= 0; j--){
            A[k][j] = A[k][j] / tmp[0];
        }
        for (int i = k + 1; i < N; i++){
            float tmp[8] ;
            for(int i=0;i<8;i++)tmp[i]=A[i][k];
            t1 = _mm256_loadu_ps(tmp);
            for (int j = N - 8; j >k;j -= 8){
                t2 = _mm256_loadu_ps(A[i] + j);
                t3 = _mm256_loadu_ps(A[k] + j);
                t4 = _mm256_sub_ps(t2,_mm256_mul_ps(t1, t3));
                _mm256_storeu_ps(A[i] + j, t4);
            }
            for (int j = k + 1; j % 8 !=(N % 8); j++){
                A[i][j] = A[i][j] - A[i][k] * A[k][j];
            }
            A[i][k] = 0;
        }
    }
}


void m_reset(){
    for(int i = 0; i < N ;i++){
        for(int j = 0;j<i;j++)
            A[i][j] = 0;
        A[i][i] = 1.0;
        for(int j = i + 1;j<N;j++)
            A[i][j] = rand();
    }
    for(int k = 0;k<N;k++)
        for(int i = k+1;i<N;i++)
            for(int j =0;j<N;j++)
                A[i][j]+=A[k][j];
}

int main()
{

    clock_t  clockBegin,clockEnd;
    m_reset();
    clockBegin = clock();
    for(int i = 0;i<count;i++)
    ori_LU();
    clockEnd = clock();
    float standard = clockEnd - clockBegin;
    cout << "ori_LU =  " << standard << "ms" << endl;

    m_reset();
    clockBegin = clock();
    for(int i = 0;i<count;i++)
    SSE_LU();
    clockEnd = clock();
    float SSE = clockEnd - clockBegin;
    cout << "SSE_LU =  " << clockEnd - clockBegin << "ms" << endl;
    cout<<"Ratio of Acc = "<<standard/SSE<<endl;

    m_reset();
    clockBegin = clock();
    for(int i = 0;i<count;i++)
    AVX_LU();
    clockEnd = clock();
    float AVX = clockEnd - clockBegin;
    cout << "AVX_LU =  " << clockEnd - clockBegin << "ms" << endl;
    cout<<"Ratio of Acc = "<<standard/AVX<<endl;
    system("pause");
    return 0;

}