为什么 C++ 只比 VBA 快 4倍？

蕾欧娜

用C++和VBA分别写了一段用随机数字测试的值的算法，用500万个随机数，发现VBA需要570毫秒，而C++则需要150毫秒。这也太让人失望了把，C++写起来那么麻烦，起码也要快个几个数量级吧？python就更别说了，居然要7秒钟。求各位大神帮我看看是不是我的代码没优化好？
（平台是Visual Studio，已经开了Release模式）

计算方法：生成,两个在0 到 之间的随机数，数一数这些数字里面有多少个落在了半径为的1/4圆的扇形里面，用这个数字代表扇形面积，用总随机数数量代表正方形面积。因为扇形的面积是 ,而正方形的面积是, 可得知=扇形面积/正方形面积 * 4
! a: h5 b7 a* C' J
8 \( M( z% {# r+ B. M; S5 ]% hC++代码：
#include "stdafx.h"#include <iostream>#include <time.h>using namespace std;void main(){        double st = clock();        double rand_max = 32767;        srand((int)time(0));        unsigned int simulate_total = 2500000;        unsigned int inside_count = 0;        unsigned int radius = rand_max * rand_max;        unsigned int randA;        unsigned int randB;        unsigned int randA_opp;        unsigned int randB_opp;        for (unsigned int i = 1; i < simulate_total; i++){                randA = rand();                randB = rand();                if ((randA * randA + randB * randB) < radius){                        inside_count++;                }                randA_opp = rand_max - randA;                randB_opp = rand_max - randB;                if ((randA_opp * randA_opp + randB_opp * randB_opp) < radius){                        inside_count++;                }                }        cout << inside_count / double(simulate_total) * 2 << endl;        cout << clock() - st << endl;}


9 m8 c# T- X0 Y- z: LVBA代码:
) R2 H3 I, k0 y1 _3 W2 f7 X9 }) E! ASub simulate_pi()Dim area_count As Double: area_count = 0Dim simulate_count As Double: simulate_count = 2500000Dim i As DoubleDim randA As Double, randB As DoubleDim randA_opp As Double, randB_opp As DoubleFor i = 1 To simulate_count    randA = Math.Rnd()    randB = Math.Rnd()    randA_opp = 1 - randA    randB_opp = 1 - randB    If randA * randA + randB * randB < 1 Then        area_count = area_count + 1    End If    If randA_opp * randA_opp + randB_opp * randB_opp < 1 Then        area_count = area_count + 1    End IfNext iDebug.Print "Estimate: ", area_count / simulate_count * 2End Sub'VBA内测时间的方法：'新建一个module，把以下代码复制进去，然后运行test.Option ExplicitPrivate Declare Function getFrequency Lib "kernel32" _Alias "QueryPerformanceFrequency" (cyFrequency As Currency) As LongPrivate Declare Function getTickCount Lib "kernel32" _Alias "QueryPerformanceCounter" (cyTickCount As Currency) As Long'Function MicroTimer() As Double'' Returns seconds.'    Dim cyTicks1 As Currency    Static cyFrequency As Currency    '    MicroTimer = 0' Get frequency.    If cyFrequency = 0 Then getFrequency cyFrequency' Get ticks.    getTickCount cyTicks1' Seconds    If cyFrequency Then MicroTimer = cyTicks1 / cyFrequencyEnd FunctionSub test()    Dim st: st = MicroTimer    Call simulate_pi    Debug.Print (MicroTimer - st) * 1000End Sub


+ i# n, u" Y, r* q. iPython代码:#!/usr/bin/python# Filename : pi_simulate.pyimport randomimport timetime_start = time.perf_counter()inside_count = 0simulate_total = 2500000randA = 0randB = 0for i in range(1, simulate_total):    randA = random.uniform(0,1)    randB = random.uniform(0,1)    if ((randA * randA + randB * randB) < 1):        inside_count = inside_count + 1    randA_opp = 1 - randA    randB_opp = 1 - randB    if ((randA_opp * randA_opp + randB_opp * randB_opp) < 1):        inside_count = inside_count + 1print (inside_count/simulate_total*2)print ("Time spent", time.perf_counter() - time_start)
8 b. x' [4 f& `  E! M9 b