Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.advocacy > #32665 > unrolled thread

Ekopath 4 Compiler Suite open sourced

Back to article view | Back to comp.os.linux.advocacy

Contents

  Ekopath 4 Compiler Suite open sourced Chris <chrisdhaag@googlemail.com> - 2011-06-13 18:36 +0000
    Re: Ekopath 4 Compiler Suite open sourced Chris Ahlstrom <ahlstromc@xzoozy.com> - 2011-06-13 15:24 -0400
      Re: Ekopath 4 Compiler Suite open sourced Chris <chrisdhaag@googlemail.com> - 2011-06-14 07:59 +0000
      Re: Ekopath 4 Compiler Suite open sourced Hadron<hadronquark@gmail.com> - 2011-06-14 10:20 +0200
        Re: Ekopath 4 Compiler Suite open sourced Chris <chrisdhaag@googlemail.com> - 2011-06-14 08:54 +0000
          Re: Ekopath 4 Compiler Suite open sourced Chris Ahlstrom <ahlstromc@xzoozy.com> - 2011-06-14 06:08 -0400
            Re: Ekopath 4 Compiler Suite open sourced William Poaster <wp@induh-vidual.net> - 2011-06-14 11:32 +0100
    Re: Ekopath 4 Compiler Suite open sourced DFS <nospam@dfs.com> - 2011-06-13 16:16 -0400
      Re: Ekopath 4 Compiler Suite open sourced Chris <chrisdhaag@googlemail.com> - 2011-06-14 08:22 +0000
        Re: Ekopath 4 Compiler Suite open sourced Hadron<hadronquark@gmail.com> - 2011-06-14 10:48 +0200
          Re: Ekopath 4 Compiler Suite open sourced Chris <chrisdhaag@googlemail.com> - 2011-06-14 16:19 +0000

#32665 — Ekopath 4 Compiler Suite open sourced

What the subject says:
http://www.pathscale.com/ekopath4-open-source-announcement

"one of the industries highest performance Intel 64 and AMD C, C++ and 
Fortran compilers."

Phoronix has some promising benchmarks:
http://www.phoronix.com/scan.php?
page=article&item=pathscale_ekopath4_open&num=3

A very good addition for the Open Source world, I'd say.

[toc] | [next] | [standalone]

#32672

Chris wrote this copyrighted missive and expects royalties:

> What the subject says:
> http://www.pathscale.com/ekopath4-open-source-announcement
>
> "one of the industries highest performance Intel 64 and AMD C, C++ and 
> Fortran compilers."
>
> Phoronix has some promising benchmarks:
> http://www.phoronix.com/scan.php?page=article&item=pathscale_ekopath4_open&num=3
>
> A very good addition for the Open Source world, I'd say.

Well, if you can get the Linux kernel and applications to compile with that
compiler, that 40% speed boost ought to silence the trolls who confuse
algorithmic speed with the speed of optimized code.


-- 
Must I hold a candle to my shames?
		-- William Shakespeare, "The Merchant of Venice"

[toc] | [prev] | [next] | [standalone]

#32800

Am Mon, 13 Jun 2011 15:24:09 -0400 schrieb Chris Ahlstrom:

> Well, if you can get the Linux kernel and applications to compile with
> that compiler,

In the phoronix forum someone said it needs one patch to do so.

> that 40% speed boost ought to silence the trolls who
> confuse algorithmic speed with the speed of optimized code.

I haven't found benchmarks from the kernel so far so I don't know how 
much faster it actually is. There seem to be huge differences in the 
speedup of different applications.

[toc] | [prev] | [next] | [standalone]

#32801

Chris Ahlstrom <ahlstromc@xzoozy.com> writes:

> Chris wrote this copyrighted missive and expects royalties:
>
>> What the subject says:
>> http://www.pathscale.com/ekopath4-open-source-announcement
>>
>> "one of the industries highest performance Intel 64 and AMD C, C++ and 
>> Fortran compilers."
>>
>> Phoronix has some promising benchmarks:
>> http://www.phoronix.com/scan.php?page=article&item=pathscale_ekopath4_open&num=3
>>
>> A very good addition for the Open Source world, I'd say.
>
> Well, if you can get the Linux kernel and applications to compile with that
> compiler, that 40% speed boost ought to silence the trolls who confuse
> algorithmic speed with the speed of optimized code.

Perhaps you can enlighten us as to which "trolls" are "confuse"
"algorithmic speed with the speed of optimized code".

You sure talk some garbage in your haste to impress people.

Algorithmic speed? Say what?

In your haste to suck up you probably missed the fact that its a BSD
license and is designed to offer speed increases for NVIdias closed tech
range too..

Unfortunately for you, and heres the keeper, the benchmarks discussed
there relate to how fast it COMPILES things. Nothing to do with the
speed of optimised code - so the 40% increase basically means jack to
most people unless you're compiling kernels daily - and even then the
compile time is but a % of the overrall build process.

Cue Creepy claiming that he meant the compiler efficiency when he said
"algorithmic efficiency" ....

*chuckle*

It's as if you cant help making an ass of yourself.

[toc] | [prev] | [next] | [standalone]

#32806

Am Tue, 14 Jun 2011 10:20:34 +0200 schrieb Hadron:

> In your haste to suck up you probably missed the fact that its a BSD
> license

In your haste you forgot to read the Phoronix article:

Page 2 below the first picture:
"PathScale is using the GNU GPL license."

But they seem to have components under all sorts of Open Source licences:
http://www.pathscale.com/pdf/PathScale_Software_License.pdf
I don't really understand from that document what exactly has which 
licence though.

> and is designed to offer speed increases for NVIdias closed tech
> range too..

What do you mean?
"ENZO is a GPGPU/multi-core solution designed for NVIDIA Tesla hardware 
with CUDA support"

But we are talking about pathcc that does "Code generation for latest AMD 
& Intel® 64 processors".

> Unfortunately for you, and heres the keeper, the benchmarks discussed
> there relate to how fast it COMPILES things. Nothing to do with the
> speed of optimised code - so the 40% increase basically means jack to
> most people unless you're compiling kernels daily - and even then the
> compile time is but a % of the overrall build process.

Phoronix is often a bit sloppy when describing what they are actually 
doing.

But the description of the last picture on page four should make it clear 
that you are probably wrong:
"In the final test of our quick benchmarks, *the TSCP chess computational 
program was 80% faster* when built with EKOPath instead of GCC."

> Cue Creepy claiming that he meant the compiler efficiency when he said
> "algorithmic efficiency" ....
> *chuckle*
> It's as if you cant help making an ass of yourself.
Someone is a bit premature here.

[toc] | [prev] | [next] | [standalone]

#32811

Chris wrote this copyrighted missive and expects royalties:

> Am Tue, 14 Jun 2011 10:20:34 +0200 schrieb Hadron:
>
> <weird uncomprehending idiocy snipped>
>
> Someone is a bit premature here.

I think it was his father.

-- 
I've read SEVEN MILLION books!!

[toc] | [prev] | [next] | [standalone]

#32815

In reply to Chris Ahlstrom who posted:

> Chris wrote this copyrighted missive and expects royalties:
>
>> Am Tue, 14 Jun 2011 10:20:34 +0200 schrieb Hadron:
>>
>> <weird uncomprehending idiocy snipped>
>>
>> Someone is a bit premature here.
>
> I think it was his father.

ROTFL!

-- 
Earth is 98% full...please delete anyone you can.
"Microsoft has vast resources, literally billions of dollars in cash, or liquid assets reserves. 
Microsoft is an incredibly successful empire built on the premise of market dominance with low-quality goods." 
-- Former White House adviser Richard A. Clarke --

[toc] | [prev] | [next] | [standalone]

#32690

On 6/13/2011 2:36 PM, Chris wrote:
> What the subject says:
> http://www.pathscale.com/ekopath4-open-source-announcement
>
> "one of the industries highest performance Intel 64 and AMD C, C++ and
> Fortran compilers."
>
> Phoronix has some promising benchmarks:
> http://www.phoronix.com/scan.php?
> page=article&item=pathscale_ekopath4_open&num=3
>
> A very good addition for the Open Source world, I'd say.


Lunix is one step closer to world domination...

Yawn!

[toc] | [prev] | [next] | [standalone]

#32802

Am Mon, 13 Jun 2011 16:16:56 -0400 schrieb DFS:

> Lunix is one step closer to world domination...

Good idea. I downloaded the Lunix source from
http://sourceforge.net/projects/lng/
and changed the cc from gcc to pathcc.

Compiles well.

So where do I get a commodore 64 to test it now?

> Yawn!

What's so boring?

[toc] | [prev] | [next] | [standalone]

#32804

Chris <chrisdhaag@googlemail.com> writes:

> Am Mon, 13 Jun 2011 16:16:56 -0400 schrieb DFS:
>
>> Lunix is one step closer to world domination...
>
> Good idea. I downloaded the Lunix source from
> http://sourceforge.net/projects/lng/
> and changed the cc from gcc to pathcc.
>
> Compiles well.
>
> So where do I get a commodore 64 to test it now?
>
>> Yawn!
>
> What's so boring?

You.

Shouldnt you be studying applied CS or something and not wasting your
time doing rote tape jockey tasks like compiling kernels you dont need?
 

[toc] | [prev] | [next] | [standalone]

#32892

Am Tue, 14 Jun 2011 10:48:09 +0200 schrieb Hadron:

> Shouldnt you be studying applied CS or something and not wasting your
> time

Yes I should. But on the other side: Who _should_ waste time?

> doing rote tape jockey tasks like compiling kernels you dont need?

It was DFS' talking about Lunix. I don't really understand why though. 
Maybe he uses a Commodore 64?

[toc] | [prev] | [standalone]

Back to top | Article view | comp.os.linux.advocacy

csiph-web