Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How does it compare to gcc?

Author: Ulrich Tuerke

Date: 03:25:25 05/17/01

Go up one level in this thread


On May 17, 2001 at 03:53:30, Roland Pfister wrote:

>On May 17, 2001 at 02:11:52, Gian-Carlo Pascutto wrote:
>
>>On May 16, 2001 at 17:41:56, Oliver Roese wrote:
>>
>>>On May 16, 2001 at 08:16:23, Gian-Carlo Pascutto wrote:
>>>
>>>>
>>>>Free trial's, but you have to fill in some forms.
>>>>
>>>>http://news.linuxprogramming.com/news_story.php3?ltsn=2001-05-15-003-06-DT-CP
>>>>
>>>>--
>>>>GCP
>>>
>>>It is about 35Mb to download.
>>>What could be expected in return for that?
>>
>>The Windows version blows both MSVC and GNU C out of the sky for
>>chessprograms...
>>
>>--
>>GCP
>
>That is a myth. It is better for *some* programs e.g. WB-Comet.
>It is worse for Patzer. Also I noticed that it does not optimize
>correctly (for Patzer).
>
>But there is a trial version, so you can test if it is usefull for you.
>
>Roland

That's true. The performance "win" from the Intel compiler also heavily depends
on the hardware architecture; on PII and Athlon, comet wins about 10% over the
MSC compiler. On the other hand, on K6 or PI, the Intel generated comet is about
10% slower than the MSC compiled executable.

Uli



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.