From: "Hibou57 (Yannick Duchêne)" <yannick_duchene@yahoo.fr>
Subject: Re: Copying rows in a two dimensional array.
Date: Sun, 14 Feb 2010 02:54:37 +0100
Date: 2010-02-14T02:54:37+01:00 [thread overview]
Message-ID: <op.u73gxbdevwnd5a@garhos> (raw)
In-Reply-To: c79a33a3-2b3e-46f1-9abf-d596f2f4b4a9@y33g2000yqb.googlegroups.com
Le Sun, 14 Feb 2010 01:42:13 +0100, jonathan <johnscpg@googlemail.com> a
écrit:
> First example: we eigen-decompose an N x N = 2048 x 2048 matrix.
> The data storage matrix is M x M = (1024+Padding) x (1024+Padding)
> Here is the running time in seconds of an iterative jacobi
> eigen-decomposition:
>
> 2048x2048: 322 sec, gnat (Padding=24)
> 2048x2048: 1646 sec, gnat (Padding=0)
> 2048x2048: 1632 sec, gfortran (Padding=0)
> 2048x2048: 1492 sec, ifort (Padding=0)
>
> The observed 500% slowdown in the absence of padded arrays is
> unacceptable, even if it is a rare event (occurring only on 2**p sized
> data arrays). In fact it's not all that rare ... more comments on
> that below. (BTW, ifort is the INTEL fortran compiler, all
> optimizations at Max. gfortran is the gcc variant.)
So this is mostly about representation clauses finally. Is that it ?
Do not know if you already know this document (as I remember I picked it
up from some one thread at comp.lang.ada), I've talked about on the other
fr.c.l.a :
http://research.scee.net/files/presentations/gcapaustralia09/Pitfalls_of_Object_Oriented_Programming_GCAP_09.pdf
I had pointed about frames #17, #18, #19 et #20, which contains good
source of inspiration. Hope this could help you to figure a path.
You've posted a long list of tests-bench and observations. I did not
looked at every thing, but hope I will have a more closer look at it later.
--
No-no, this isn't an oops ...or I hope (TM) - Don't blame me... I'm just
not lucky
next prev parent reply other threads:[~2010-02-14 1:54 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-02-01 2:11 Copying rows in a two dimensional array Peter C. Chapin
2010-02-01 4:42 ` Jeffrey R. Carter
2010-02-01 6:55 ` Niklas Holsti
2010-02-01 23:36 ` Peter C. Chapin
2010-02-04 4:27 ` Hibou57 (Yannick Duchêne)
2010-02-01 8:37 ` Dmitry A. Kazakov
2010-02-02 0:11 ` Randy Brukardt
2010-02-07 16:13 ` Robert A Duff
2010-02-08 6:30 ` tmoran
2010-02-08 13:15 ` Robert A Duff
2010-02-08 13:45 ` Dmitry A. Kazakov
2010-02-08 21:20 ` Robert A Duff
2010-02-08 23:26 ` (see below)
2010-02-09 0:36 ` Randy Brukardt
2010-02-09 1:03 ` (see below)
2010-02-09 7:11 ` Pascal Obry
2010-02-09 8:14 ` AdaMagica
2010-02-09 14:33 ` Robert A Duff
2010-02-09 1:05 ` Adam Beneschan
2010-02-09 14:45 ` Robert A Duff
2010-02-09 18:50 ` tmoran
2010-02-09 19:51 ` Pascal Obry
2010-02-09 23:03 ` Robert A Duff
2010-02-08 18:53 ` tmoran
2010-02-08 21:14 ` Robert A Duff
2010-02-08 21:29 ` Pascal Obry
2010-02-09 8:56 ` Jean-Pierre Rosen
2010-02-09 9:14 ` AdaMagica
2010-02-09 11:19 ` Jean-Pierre Rosen
2010-02-09 14:26 ` Robert A Duff
2010-02-09 6:34 ` tmoran
2010-02-09 14:29 ` Robert A Duff
2010-02-09 18:49 ` tmoran
2010-02-09 22:58 ` Robert A Duff
2010-02-01 22:10 ` Jerry
2010-02-02 0:07 ` Randy Brukardt
2010-02-02 8:52 ` Jean-Pierre Rosen
2010-02-02 22:23 ` Jerry
2010-02-03 1:24 ` Adam Beneschan
2010-02-04 4:42 ` Hibou57 (Yannick Duchêne)
2010-02-14 0:42 ` jonathan
2010-02-14 1:54 ` Hibou57 (Yannick Duchêne) [this message]
2010-02-14 16:16 ` jonathan
2010-03-22 8:56 ` Ole-Hjalmar Kristensen
2010-02-16 6:51 ` David Thompson
2010-02-04 4:13 ` Hibou57 (Yannick Duchêne)
2010-02-04 9:10 ` Dmitry A. Kazakov
2010-02-04 9:23 ` Hibou57 (Yannick Duchêne)
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox