From: Lutz Donnerhacke <lutz@iks-jena.de>
Subject: Re: fastest data structure
Date: Wed, 11 Dec 2002 15:11:06 +0000 (UTC)
Date: 2002-12-11T15:11:06+00:00 [thread overview]
Message-ID: <slrnavelc6.jk.lutz@taranis.iks-jena.de> (raw)
In-Reply-To: at7k2o$9fb$1@news-reader10.wanadoo.fr
* Etienne Baudin wrote:
> The process is the same with the 3 test (always reading "un") , and the
> running time seems depend on the record size....??
The increment can be done in a simple (incrementing by 4) or complex
(increment by 17) way. Processors are optimized on special increments. The
other way to compute the new address is even optimized on special offsets.
Futhermore smaller records increases the data locality, so the cache does a
better job on smaller data.
next prev parent reply other threads:[~2002-12-11 15:11 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-12-10 21:01 fastest data structure Etienne Baudin
2002-12-10 22:12 ` Victor Porton
2002-12-11 1:14 ` Jeffrey Carter
2002-12-11 3:23 ` SteveD
2002-12-11 13:03 ` Marin David Condic
2002-12-11 15:02 ` Etienne Baudin
2002-12-11 15:11 ` Lutz Donnerhacke [this message]
2002-12-11 19:04 ` tmoran
2002-12-12 4:22 ` SteveD
2002-12-12 12:40 ` P R Keeble
2002-12-14 16:23 ` Simon Wright
2002-12-17 0:33 ` Randy Brukardt
-- strict thread matches above, loose matches on Subject: below --
2002-12-10 21:51 Gautier direct_replies_not_read
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox