News for hackers. Stuff that matters.
Re:Playing with Go1.1beta2, it's much faster by dscrd on Tuesday April 09, 2013 @01:20AM (#5516560)
According to the Shootout, Go-1.1 is only slower on a few tests than Java, while being a bit faster on most and while typically using way less memory.
Re:Playing with Go1.1beta2, it's much faster by voidlogic on Tuesday April 09, 2013 @01:25AM (#5516585)
Win some, lose some: http://benchmarksgame.alioth.debian.org/u64/benchmark.php?te...

Note: The much faster version Go's DNA regex that uses PCRE rather than "regexp", but it is currently broken due to a build error. (http://benchmarksgame.alioth.debian.org/u64/program.php?test...)

Re:Playing with Go1.1beta2, it's much faster by igouy on Tuesday April 09, 2013 @10:29AM (#5518252)
Let's also note that the Go regex-dna programs run out of memory on x86 :-(
Re:Playing with Go1.1beta2, it's much faster by voidlogic on Tuesday April 09, 2013 @11:12AM (#5518620)
Re:Playing with Go1.1beta2, it's much faster by igouy on Tuesday April 09, 2013 @11:22AM (#5518704)
Thanks, let's hope that holds for the Go #7 program.
Re:Playing with Go1.1beta2, it's much faster by trailfox on Tuesday April 09, 2013 @02:59AM (#5516809)
The quad core results indicate the exact opposite in terms of performance with the JVM clearly ahead: http://benchmarksgame.alioth.debian.org/u32q/benchmark.php?t...

Go uses less memory than Java 7, but then again so does Java 1.1

Re:Playing with Go1.1beta2, it's much faster by gillianseed on Tuesday April 09, 2013 @09:07AM (#5517774)
Go does slightly better in the 64-bit test (which I think is more apt to link to rather than the 32-bit which you keep using)

http://benchmarksgame.alioth.debian.org/u64q/benchmark.php?t...

That said, Java still beats Go quite readily in this benchmark spread, particularly in regex and binary-trees where Go does extremely poor. It should be noted that lots of languages implement their regex libraries in C for speed, while Go's implementation is written in Go. Still, I think it should do better than this.

That Java beats Go overall is not particularly surprising, Java is very mature and it's jit-compiler has very advanced optimizations, Perhaps by Go 2.0 we'll see where Go 'lands' in terms of performance, there's obviously alot more to be had.

Re:Playing with Go1.1beta2, it's much faster by voidlogic on Tuesday April 09, 2013 @09:52AM (#5518018)
The u64q link you posted is still using Go 1.0.3, you can see some 1.1 beta 1 results here: http://benchmarksgame.alioth.debian.org/u64/benchmark.php?te...
Re:Playing with Go1.1beta2, it's much faster by trailfox on Tuesday April 09, 2013 @10:46AM (#5518397)
Thanks, the 32 bit results on quad core are Go 1.1 beta:

http://benchmarksgame.alioth.debian.org/u32q/benchmark.php?t...

Re:Playing with Go1.1beta2, it's much faster by voidlogic on Tuesday April 09, 2013 @10:51AM (#5518427)
Go does has known issues and does poorly on 32-bit x86, but that is OK, no one uses it :)-
Re:Playing with Go1.1beta2, it's much faster by gillianseed on Tuesday April 09, 2013 @10:52AM (#5518441)
Thanks! Are there any 1.1 beta 1 results for Go with quad core rather than one core?
Re:Playing with Go1.1beta2, it's much faster by carbocation on Tuesday April 09, 2013 @08:13AM (#5517560)
Reading some of the go programs, they do not appear to be configured to take advantage of more than one core.
Re:Playing with Go1.1beta2, it's much faster by igouy on Tuesday April 09, 2013 @10:25AM (#5518218)
Please -- no generalities -- provide a URL so we can all see what you are looking at, and reach our own understanding.
Re:Playing with Go1.1beta2, it's much faster by carbocation on Tuesday April 09, 2013 @03:37PM (#5520636)
OK, here's the first program on the list provided by trailfox:

http://benchmarksgame.alioth.debian.org/u32q/program.php?tes...

Note that:

1) It only ever uses one core, because

2) It isn't configured with GOMAXPROCS, which therefore defaults to 1.

So, sure, they have an N-core machine, but they're only allowing go to access 1 core.

Re:Playing with Go1.1beta2, it's much faster by igouy on Tuesday April 09, 2013 @06:48PM (#5521963)
>>but they're only allowing<<

Notice on the list provided by `trailfox` the "≈ CPU Load" column -- when the programs are written to use 4 cores "they're" allowing Go to access 4 cores.

Re:Playing with Go1.1beta2, it's much faster by carbocation on Wednesday April 10, 2013 @11:52AM (#5525877)
No. Go programs currently require a one-line instruction to tell them how many cores to use. Without that one line, they will only use one core.

The go program in question was written without that instruction; therefore, it's only allowed to use one core.

Re:Playing with Go1.1beta2, it's much faster by igouy on Wednesday April 10, 2013 @12:47PM (#5526310)
>only allowed to use one core<

Only "allowed" by who?

If you mean that "the Go program in question" is by-design a sequential program then -- yes of course it is!

The contributed Go programs that were written for multi-core use more than one core -- http://benchmarksgame.alioth.debian.org/u32q/program.php?tes...

Re:Playing with Go1.1beta2, it's much faster by Roboprog on Tuesday April 09, 2013 @09:37AM (#5517931)
What, you don't need a mandlebrot set in your CRUD app?