Sebastien is right! Thanks a lot! I forgot that the optimizer does
eliminate some code at times. (and forgot
this https://dave.cheney.net/2013/06/30/how-to-write-benchmarks-in-go)
This is the C code now (compiled with gcc -O2 run_std_pow.c -o run_std_pow
-lm):
#include "stdio.h"
#include "math.h"
int main() {
double res = 0.0;
double x = 2.5;
int Nmax = 10000000;
for (int N=0; N<Nmax; N++) {
for (int i=0; i<20; i++) {
res += pow(x, i);
}
}
printf("res = %g\n", res);
}
and this is the Go code:
package main
import (
"fmt"
"math"
)
func main() {
res := 0.0
x := 2.5
Nmax := 10000000
for N := 0; N < Nmax; N++ {
for i := 0; i < 20; i++ {
res += math.Pow(x, float64(i))
}
}
fmt.Printf("res = %g\n", res)
}
The C code runs on 8.262s and the Go code on 5.414s!
Cheers.
D
On Friday, August 4, 2017 at 5:39:07 PM UTC+10, Sebastien Binet wrote:
>
> Dorival,
>
> On Fri, Aug 4, 2017 at 8:20 AM, Dorival Pedroso <dorival...@gmail.com
> <javascript:>> wrote:
>
>> I've noticed that this C code:
>>
>> #include "math.h"
>> int main() {
>> double x = 2.5;
>> int Nmax = 10000000;
>> for (int N=0; N<Nmax; N++) {
>> for (int i=0; i<20; i++) {
>> pow(x, i);
>> }
>> }
>> }
>>
>> can run up to 50x faster than this Go code:
>>
>> package main
>>
>> import "math"
>>
>> func main() {
>> x := 2.5
>> Nmax := 10000000
>> for N := 0; N < Nmax; N++ {
>> for i := 0; i < 20; i++ {
>> math.Pow(x, float64(i))
>> }
>> }
>> }
>>
>> The C code was compiled with: gcc -O2 ccode.c -o ccode -lm
>> then run with time ./ccode
>>
>> The Go code was compiled with: go build gcode.go
>> then run with time ./gcode
>>
>> I've used the time command on Linux (Ubuntu) to get some estimate.
>>
>> So the question is: how can we make the Go code faster?
>>
> no, the question is whether you are really doing what you think you are
> doing in C.
> that's the usual question with benchmarking.
>
> with your initial C code, I get this on my machine:
>
> $> time ./c-code
> real 0m0.261s
> user 0m0.260s
> sys 0m0.000s
>
> and this for the Go code:
>
> $> time ./go-code
> real 0m12.781s
> user 0m12.784s
> sys 0m0.000s
>
> then, modifying the C code to make sure the compiler won't elide the pow
> computation altogether:
>
> #include <math.h>
> #include <stdio.h>
>
> int main() {
> double sum = 0;
> double x = 2.5;
> int Nmax = 10000000;
> for (int N=0; N<Nmax; N++) {
> for (int i=0; i<20; i++) {
> sum += pow(x, i);
> }
> }
> printf("sum=%f\n", sum);
> }
>
> I now get:
>
> $> time ./c-code-sum
> sum=606329794183272.375000
>
> real 0m18.385s
> user 0m18.377s
> sys 0m0.000s
>
> applying the same modifications to the Go code:
>
> $> time ./go-code-sum
>
> sum=6.063297941832724e+14
>
> real 0m12.685s
> user 0m12.673s
> sys 0m0.010s
>
> -s
>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "golang-nuts" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to golang-nuts...@googlegroups.com <javascript:>.
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
