tensorflow GPU小测试
2019.01.18补充:这个例子不具有代表性,涉及到卷积运算的时候,GPU的加速效果会体现得比较明显。
简单测试了一下tensorflow的GPU计算和CPU计算的区别。这里的计算例子只非常简单的小规模矩阵相乘,但是也体现出了CPU和GPU算力的差距,代码及结果如下:
import tensorflow as tf
import datetime
#running
# Creates a graph.(cpu version)
print('cpu version')
starttime1 = datetime.datetime.now()
with tf.device('/gpu:0'):
a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0,1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[6, 9], name='a')
b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0,1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[9, 6], name='b')
c = tf.matmul(a, b)
c = tf.matmul(c,a)
c = tf.matmul(c,b)
# Creates a session with log_device_placement set to True.
sess1 = tf.Session(config=tf.ConfigProto(log_device_placement=True))
# Runs the op.
for i in range(59999):
sess1.run(c)
print(sess1.run(c))
sess1.close()
endtime1 = datetime.datetime.now()
time1 = (endtime1 - starttime1).microseconds
#print('time1:',time1)
#############################################
print('gpuversion')
# Creates a graph.(gpu version)
starttime2 = datetime.datetime.now()
#running
with tf.device('/gpu:0'):
a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0,1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[6, 9], name='a')
b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0,1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[9, 6], name='b')
c = tf.matmul(a, b)
c = tf.matmul(c,a)
c = tf.matmul(c,b)
# Creates a session with log_device_placement set to True.
sess2 = tf.Session(config=tf.ConfigProto(log_device_placement=True))
# Runs the op.
for i in range(59999):
sess2.run(c)
print(sess2.run(c))
sess2.close()
endtime2 = datetime.datetime.now()
time2 = (endtime2 - starttime2).microseconds
print('time1:',time1)
print('time2:',time2)
结果如下:
cpu version
[[
18225.
36450.
54675.
72900.
91125.
109350.]
[
24300.
48600.
72900.
97200.
121500.
145800.]
[
18225.
36450.
54675.
72900.
91125.
109350.]
[
24300.
48600.
72900.
97200.
121500.
145800.]
[
18225.
36450.
54675.
72900.
91125.
109350.]
[
24300.
48600.
72900.
97200.
121500.
145800.]]
gpuversion
[[
18225.
36450.
54675.
72900.
91125.
109350.]
[
24300.
48600.
72900.
97200.
121500.
145800.]
[
18225.
36450.
54675.
72900.
91125.
109350.]
[
24300.
48600.
72900.
97200.
121500.
145800.]
[
18225.
36450.
54675.
72900.
91125.
109350.]
[
24300.
48600.
72900.
97200.
121500.
145800.]]
time1: 356158
time2: 363249
注:以上时间单位是微秒
最后
以上就是受伤鼠标最近收集整理的关于tensorflow GPU小测试的全部内容,更多相关tensorflow内容请搜索靠谱客的其他文章。
本图文内容来源于网友提供,作为学习参考使用,或来自网络收集整理,版权属于原作者所有。
发表评论 取消回复