AVX向量化学习(五)-INT型数组相加操作
使用AVX指令集进行2个INT型的数组相加操作
使用到的AVX函数介绍
1.
| __m256i _mm256_loadu_si256 (__m256i const * mem_addr)
|
Synopsis
m256i _mm256_loadu_si256 (m256i const * mem_addr)
#include <immintrin.h>
Instruction: vmovdqu ymm, m256
CPUID Flags: AVX
Description
Load 256-bits of integer data from memory into dst. mem_addr does not need to be aligned on any particular boundary.
Operation
| dst[255:0] := MEM[mem_addr+255:mem_addr]
dst[MAX:256] := 0
|
Performance
| Architecture | Latency | Throughput (CPI) |
|---|
| Icelake | 7 | 0.5 |
| Skylake | 7 | 0.5 |
| Broadwell | 1 | 0.25 |
| Haswell | 1 | 0.25 |
| Ivy Bridge | 1 | 0.5 |
2.
| __m256i _mm256_add_epi32 (__m256i a, __m256i b)
|
Synopsis
m256i _mm256_add_epi32 (m256i a, __m256i b)
#include <immintrin.h>
Instruction: vpaddd ymm, ymm, ymm
CPUID Flags: AVX2
Description
Add packed 32-bit integers in a and b, and store the results in dst.
Operation
| FOR j := 0 to 7
i := j*32
dst[i+31:i] := a[i+31:i] + b[i+31:i]
ENDFOR
dst[MAX:256] := 0
|
Performance
| Architecture | Latency | Throughput (CPI) |
|---|
| Icelake | 1 | 0.33 |
| Skylake | 1 | 0.33 |
| Broadwell | 1 | 0.5 |
| Haswell | 1 | 0.5 |
程序源代码
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
| #include<stdio.h>
#include <immintrin.h>
int main()
{
int a[40];
int b[40];
int i=0;
int ans1[40];
int ans2[40];
for (i=0;i<40;i++)
{
a[i]=i;
b[i]=2*i;
}
for (i=0;i<40;i++)
{
ans1[i]=a[i]+b[i];
}
printf("串行计算结果：\n");
for (i=0;i<40;i++)
{
printf("%d ",ans1[i]);
}
printf("\n");
__m256i v0;
__m256i v1;
__m256i v2;
for (i=0;i<40-8;i+=8)
{
v0 = _mm256_loadu_si256((const __m256i*)(a+i));
v1 = _mm256_loadu_si256((const __m256i*)(b+i));
v2 = _mm256_add_epi32(v0,v1);
_mm256_storeu_si256((__m256i*)(ans2+i),v2);
}
for (;i<40;i++)
{
ans2[i]=a[i]+b[i];
}
printf("并行计算结果：\n");
for (i=0;i<40;i++)
{
printf("%d ",ans2[i]);
}
return 0;
}
|
程序输出
| 串行计算结果：
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 96 99 102 105 108 111 114 117
并行计算结果：
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69 72 75 78 81 84 87 90 93 96 99 102 105 108 111 114 117
|
相关链接
[https://software.intel.com/sites/landingpage/IntrinsicsGuide/]: “ Intel® Intrinsics Guide”