比较c++中unordered_map和数组的性能

当有一个100万的kv数据，并且k是小于100万的数字，是该采用unordered_map还是数组来存储呢；结论是采用数组来访问，数组性能比unordered_map快35倍

#include <unordered_map>
#include <sys/time.h>
#include <iostream>

using namespace std;
int main() {
    std::unordered_map<int, int> map;
    int* a = new int[1000];
    int* b = new int[1000];
    for (int i=0; i< 1000; ++i) {
        map[i] = i;
        a[i] = i;
        b[i] = (i*11 + i*i)%1000;
    }
    int sum = 0;

    struct timeval t1,t2;
    double timeuse;
    gettimeofday(&t1,NULL);

    for (int i = 0; i < 10000; ++i)
        for (int j = 0; j < 1000; ++j) {
            sum += map[j];
        }

    gettimeofday(&t2,NULL);
    timeuse = (t2.tv_sec - t1.tv_sec) + (double)(t2.tv_usec - t1.tv_usec)/1000000.0;
    cout<<"time = "<<timeuse<<endl;
    sum = 0;

    gettimeofday(&t1,NULL);
    for (int i = 0; i < 10000; ++i)
        for (int j = 0; j < 1000; ++j) {
            sum += a[j];
        }
    gettimeofday(&t2,NULL);
    timeuse = (t2.tv_sec - t1.tv_sec) + (double)(t2.tv_usec - t1.tv_usec)/1000000.0;
    cout<<"time = "<<timeuse << " " << map.size() <<endl;

    return 0;
}

输出

time = 0.758736  // unordered_map
time = 0.026855  // 数组