java 多线程并行求数值积分（πPI） 之 join() 方法应用

现在 先看一个 使用Runnable接口实现求数值积分的java多线程并行程序如下（电脑是四核，并行3个线程）：

package com.xing.add;
/**
 * 
 * @author Yinxing
 *
 */
public class PI_Runnable {

	public static void main(String[] args) throws InterruptedException {
		// TODO Auto-generated method stub
		//-------------------------并行运算------------------------------
		long startTime ,endTime;
		double step = 1.0 / 100000000;
		works works1 = new works(1, 100000000);
		works works2 = new works(2, 100000000);
		works works3 = new works(3, 100000000);
		Thread thread1 = new Thread(works1);
		Thread thread2 = new Thread(works2);
		Thread thread3 = new Thread(works3);
		startTime = System.currentTimeMillis();//获取并行计算的开始时的系统时间
		thread1.start();
		thread2.start();
		thread3.start();
		thread1.join();//实现线程并行
		thread2.join();
		thread3.join();
		endTime = System.currentTimeMillis();//获取并行计算的结束时的系统时间
		long ParallelTime = endTime - startTime;
		System.out.println("PI(π) = "+ (works1.getSum() + works2.getSum() + works3.getSum()) * step);//并行计算求的PI(π)值
		System.out.println("Runnable求PI(π)并行时间："+ ParallelTime);
		//---------------------------串行运算---------------------------------
		works works5 = new works(1, 100000000);
		startTime = System.currentTimeMillis();//获取串行计算的开始时的系统时间
		double PI = works5.GetPI();
		endTime = System.currentTimeMillis();//获取串行计算的结束时的系统时间
		long SeriaTime = endTime - startTime;
		System.out.println("PI(π) = "+ PI * step);//串行计算求的PI(π)值
		System.out.println("求PI(π)串行时间："+ SeriaTime);
		//------------------------------输出相对加速比-------------------------
		System.out.println("相对加速比: "+SeriaTime +"/"+ParallelTime);
	}
}

class works implements Runnable{

	private long start;
	private long end ;
	private double sum;
	private double step;
	private long num_steps = 100000000;
	public works(long start,long end){
		super();
		this.start = start;
		this.end = end;
	}
	@Override
	public void run() {
		// TODO Auto-generated method stub
		double x = 0;
		step = 1.0 / (double) num_steps;
		for (long i = start; i <= end; i += 3) {
			x = (i + 0.5) * step;
			sum = sum + 4.0 / (1.0 + (x*x));
		}
	}
	public double GetPI(){
		double x = 0;
		step = 1.0 / (double) num_steps;
		for (long i = start; i <= end; i ++) {
			x = (i + 0.5) * step;
			sum = sum + 4.0 / (1.0 + (x*x));
		}
		return sum;
	}
	public double getSum() {
		return sum;
	}
}

运行结果如下：



我们可以看出线程

thread1.start();
		thread2.start();
		thread3.start();

在调用join()方法

thread1.join();//实现线程并行
		thread2.join();
		thread3.join();

使其同时在多核上并行处理。由于其他的因素 加速比 1350/460=2.9347826086956 虽然不等于3，但是并行的效果是显而易见的