关于python多线程效率

在网上看到python多线程

今天尝试一下，先贴源码

import cv2
from time import ctime


def music(func):
    for i in range(100):
        img = cv2.imread(fn, 2)
      #  print "I was at the %s! %s" %(func,ctime())

def move(func):
    for i in range(100):
        img2 = cv2.imread(fn2, 2)
      #  print "I was at the %s! %s" %(func,ctime())

threads = []
t1 = threading.Thread(target=music,args=(u'imread1',))
threads.append(t1)
t2 = threading.Thread(target=move,args=(u'forshow1',))
threads.append(t2)

if __name__ == '__main__':
    try: fn = sys.argv[1]
    except:fn = 'E:\\5.bmp'
    fn2 ='E:\\6.bmp'
    print "1 was start at the %s" %(ctime())
    for t in threads:
        t.setDaemon(True)
        t.start()
    t.join()
    print "1 was end at the %s" %(ctime())
    print "2 was start at the %s" %(ctime())
    music(u'imread2')
    move(u'imread2')
    print "2 was end at the %s" %(ctime())
    print "all over %s" %ctime()

最后结果如下图，可以发现多线程效率提高还是有的，真是线程本身的开销也较大，多开一个线程其效率提高在30%左右。并不像网上所说的是鸡肋，原因还在探究，但是windows对线程有自动优化，现在还不清楚，有待继续深入