tbb库的sort多线程优化

tbb库的sort多线程优化

使用ubuntu20，安装tbb库

sudo apt update
sudo apt install libtbb-dev

测试代码

#include <iostream>
#include <algorithm>
#include <chrono>
#include <random>
#include <ratio>
#include <vector>
#include <execution>
#include <tbb/blocked_range.h>
using namespace std;
int main()
{
	random_device rd;
	vector<double> vals(1000000);
	for(auto& d:vals)
	{
		d = static_cast<double>(rd());
	}

	for(int i=0;i<5;++i)
	{
		using namespace std::chrono;
		vector<double> sorted(vals);
		const auto startTime = high_resolution_clock::now();
		sort(std::execution::par, sorted.begin(), sorted.end());
		//sort(sorted.begin(), sorted.end());
		const auto endTime = high_resolution_clock::now();
		cout<< "Latency:"
			<<duration_cast<duration<double,milli>>(endTime-startTime).count()
			<<endl;
	}
}

编译

g++ test.cpp  -std=c++2a -ltbb

不使用多线程优化结果（使用恢复注释代码）

Latency:352.229
Latency:352.568
Latency:352.57
Latency:353.675
Latency:355.484

使用多线程优化结果

Latency:126.013
Latency:113.639
Latency:112.808
Latency:112.826
Latency:112.7