multithreading - Basic Threading in C++ Program Issue -

#include <stdlib.h> #include <iostream> #include <thread>  using namespace std; int threshold = 20;  __m128i set1 = _mm_set1_epi8(1); __m128i set0 = _mm_set1_epi8(0); __m128i set2 = _mm_set1_epi8(2);  #define capture_perf(x) x = capturetime(); static unsigned long capturetime() {        double curtime = 1000 * (double)clock() / clocks_per_sec;        return (unsigned long)curtime; }  void threadfunc1(unsigned char *threshold_tab ,int start , int end ) {     (int = start;i < end;i += 16)         _mm_storeu_si128((__m128i *)(threshold_tab + + 255), set1); } void threadfunc2(unsigned char *threshold_tab ,int start , int end ) {     (int = start;i <= end;i += 16)         _mm_storeu_si128((__m128i *)(threshold_tab + + 255), set0); } void threadfunc3(unsigned char *threshold_tab ,int start , int end ) {     (int = start;i <= end;i += 16)             _mm_storeu_si128((__m128i *)(threshold_tab + + 255), set2); }  int main() {     int i;     unsigned long int start, end;     unsigned char  threshold_tab[530];          __m128i set1 = _mm_set1_epi8(1);         __m128i set0 = _mm_set1_epi8(0);         __m128i set2 = _mm_set1_epi8(2);          capture_perf(start);         for(int j = 0 ; j < 10000; j++) {             thread thread1(threadfunc1, threshold_tab, -255,-threshold);             thread1.join();         };             capture_perf(end);         unsigned long int time = (end - start);         cout << " time taken :"<<  time <<"msec"<< endl;         return 0; } 

there 2 pieces of code runs without threading , and threading. system 4 core systems. ran 10000 times see performance improvement of threading. without threading ran in 1msec threading ran 27msec. why ? newbie threading . please me on .