When starting a new thread, there will be additional overhead for a certain period of time, so not all cases should use multithreading. The following code snippet is used for a simple experiment:
// TimeCostToStartNewThread. javapublic class TimeCostToStartNewThread {public static void main (String [] args) {long now1 = System. nanoTime (); for (int I = 0; I <10000000; I ++) {// I values:,} long current1 = System. nanoTime (); System. out. println ("current-now 1 =" + (current1-now1); long now2 = System. nanoTime (); new Thread (new Runnable () {public void run () {for (int I = 0; I <10000000; I ++ ){}}}). start (); long current2 = System. nanoTime (); System. out. println ("current-now 2 =" + (current2-now2 ));}}
From the above results, we can know that the additional overhead for starting a new thread is about 4*(6 to the power of 10) nanoseconds. When the overhead time for using sequential programming is greater than this value, we can consider multithreading.
Start building with 50+ products and up to 12 months usage for Elastic Compute Service
1 on 1 presale consultation
24/7 Technical Support 6 Free Tickets per Quarter Faster Response
Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.
A comprehensive suite of global cloud computing services to power your business
Payment Methods We Support
