Which one is faster to process between Math.pow() and using * operator several times?

Question

Am writing a java program that raises different numbers to powers. Between Math.pow (8,4) and 8*8*8*8, which one takes shorter time to be processed?

Have you already tried to perform both computations, let's say 1,000,000 times using a `for` loop? (just to compare execution times) — xav, May 11 '14 at 17:17
@xav I've not tried using big numbers. Lemme try and get back to you — Aroniez, May 11 '14 at 17:24

score 0 · Answer 1 · edited May 11 '14 at 18:59

0

Efficiency of Power function is greater than loop.You can check time taken for each functionility in following manner:-

    double i=10;
    java.util.Date dt=new java.util.Date();
    double z=(long) Math.pow (8,i);
    java.util.Date dt1=new java.util.Date();
    System.out.println("Using Power formula:"+(dt1.getTime()-dt.getTime()));
    double t=1;
    for(double k=1;k<=i;k++){
        t*=8;
    }
    java.util.Date dt2=new java.util.Date();
    System.out.println("Using loop :"+(dt2.getTime()-dt1.getTime()));

edited May 11 '14 at 18:59

Nidhish Krishnan

20,593
6
63
76

answered May 11 '14 at 18:23

Subhajit Majumder

9
2

You can't measure such a fast operation with Date... See also http://stackoverflow.com/questions/504103/how-do-i-write-a-correct-micro-benchmark-in-java – assylias May 12 '14 at 07:20

score 0 · Answer 2 · answered May 13 '14 at 09:55

For simple operations like 8^4, a manual multiplication will be faster. A micro benchmark using jmh shows (nanoseconds/op):

Benchmark                 Mode Thr    Cnt  Sec         Mean   Mean error    Units
c.a.p.SO23595573.mult     avgt   1      3    2        3.074        0.022  nsec/op
c.a.p.SO23595573.pow      avgt   1      3    2       68.696        0.186  nsec/op

The code:

private int x = 8;
private int y = 4;

@GenerateMicroBenchmark
public double mult() {
    double result = x;
    for (int i = 0; i < y - 1; i++) {
        result *= x;
    }
    return result;
}

@GenerateMicroBenchmark
public double pow() {
    return Math.pow(x, y);
}

score -1 · Answer 3 · answered May 11 '14 at 17:17

-1

Both the functions will be calculated in compile-time and would have the same efficiency.

answered May 11 '14 at 17:17

Dmitry Ginzburg

7,391
2
37
48

Is there a way i can check the time taken for a function to be calculated? – Aroniez May 11 '14 at 17:28
That's another question. What have you tried to get it? – Dmitry Ginzburg May 11 '14 at 17:33
I've tried using big numbers and a timer but i can't observe a noticeable difference. I just to confirm the difference – Aroniez May 11 '14 at 17:36
'course, you won't get it while making operations with one or two numbers. Make a bit data set: 1_000_000, for example – Dmitry Ginzburg May 11 '14 at 17:44
Lemme try that. Thanks – Aroniez May 11 '14 at 17:47
@ginz `Math.pow()` won't be evaluated at compile time and `8*8*8` may or may not be evaulated at compile time depending on how it is written in the code. – assylias May 11 '14 at 17:52

score -1 · Answer 4 · edited May 11 '14 at 19:02

-1

Check this small code

long pr=1;
long st=System.currentTimeMillis();
for(int i=0;i<200_00_00_000;i++)
   pr=pr*8;
long en=System.currentTimeMillis();
System.out.println("Executiom time  : "+(en-st));

It'll give you execution time for this code.
Do the same for Math.pow(8,200_00_00_000) and find the difference yourself.

P.S : Math.pow() executes faster.

edited May 11 '14 at 19:02

Nidhish Krishnan

20,593
6
63
76

answered May 11 '14 at 17:57

amanraj

47
1
6

-1 Your measurement is flawed (no warmup, you don't use the result of the operation etc.). see also http://stackoverflow.com/questions/504103/how-do-i-write-a-correct-micro-benchmark-in-java – assylias May 12 '14 at 07:20

Which one is faster to process between Math.pow() and using * operator several times?

4 Answers4