Which is faster in java

Question

Which one is faster in Java?

a. for(int i = 100000; i > 0; i--) {}

b. for(int i = 1; i < 100001; i++) {}
Which one is faster in Java?

a. Math.max(a,b);

b. (a>b)?a:b

I would imagine the difference would be negligible and would vary from environment to environment. — Tyler Treat, Jan 11 '11 at 01:30
The first two take the same amount of time - 0 - using my java compiler. — President James K. Polk, Jan 11 '11 at 01:32
This has been asked a million times before with some good explanations (some, however, are now just historical references). In this case, since it's homework, why not 1) try to answer it yourself 2) justify whatever answer you come up with. — , Jan 11 '11 at 01:33
since 1.a and 1.b have no side effects, they may be completely elided. 2.a and 2.b may be equivalent if hotspot decides to inline the method call. — Ron, Jan 11 '11 at 01:54

score 15 · Accepted Answer · edited May 23 '17 at 12:04

15

When you want to know what is faster, time it.

If you want to know why something is faster, that's different entirely.

edited May 23 '17 at 12:04

Community

1
1

answered Jan 11 '11 at 01:28

Daniel DiPaolo

55,313
14
116
115

wiser words are rarely spoken :) – Mihai Toader Jan 11 '11 at 01:32
1

True enough in non-trivial cases, but you have to ask yourself, are you executing these operations often enough that minuscule differences between them add up? Does it matter if one operation is 1 ns faster than another? – Chinmay Kanchi Jan 11 '11 at 01:37

score 3 · Answer 2 · answered Jan 11 '11 at 01:29

3

Run them both and test them. The difference is going to very small and possibly different on each environment.

answered Jan 11 '11 at 01:29

jzd

23,473
9
54
76

score 3 · Answer 3 · answered Jan 11 '11 at 06:53

It is worth remembering that the JVM can compile code which does nothing down to nothing, making the difference about how and when the JVM optimises the code rather than which code is best.

The real question you should be asking yourself is; Why code is clearer to understand? That is what you should use.

score 2 · Answer 4 · answered Jan 11 '11 at 01:30

2

micro benchmark them using Google Caliper

answered Jan 11 '11 at 01:30

Aravind Yarram

78,777
46
231
327

score 1 · Answer 5 · answered Jan 11 '11 at 05:13

1

Test both two answers and time them. I doubt that there will be any significant difference.

answered Jan 11 '11 at 05:13

Kevin Ji

10,479
4
40
63

score 0 · Answer 6 · answered Jan 11 '11 at 01:34

In these cases, it just doesn't matter. All these operations are going to have such negligible running time compared to the VM and GC overhead that you won't save more than a few cycles in total. Even if the difference between the two was, say a 1000 cycles, you're still talking about a difference of 1 microsecond on a 1 GHz processor. It just doesn't matter

Which is faster in java

6 Answers6