caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Gerd Stolpmann <info@gerd-stolpmann.de>
To: Jon Harrop <jonathandeanharrop@googlemail.com>
Cc: "'Martin Jambon'" <martin.jambon@ens-lyon.org>,
	Caml List <caml-list@inria.fr>
Subject: RE: [Caml-list] Efficient OCaml multicore -- roadmap?
Date: Wed, 20 Apr 2011 22:30:53 +0200	[thread overview]
Message-ID: <1303331453.8429.1282.camel@thinkpad> (raw)
In-Reply-To: <029701cbff90$7a874510$6f95cf30$@ffconsultancy.com>

Am Mittwoch, den 20.04.2011, 20:23 +0100 schrieb Jon Harrop:
> Martin wrote:
> > On 03/25/11 08:44, Hugo Ferreira replied:
> > > I'll stick to my guns here. It simply makes solving certain problem
> > > unfeasible. Point in case: I work on machine learning algorithms. I
> > > use large data-structures that must be processed (altered) in order to
> > > learn. Because these data-structures are large it become impractical
> > > to copy this to a process every time I start off a new "thread".
> > 
> > The solution would be to use get/set via a message-passing interface.
> 
> The main objective when parallelizing a program is to remove focal points because they are sources of contention. Funnelling all IO through a single getter/setter would add a focal point and, most likely, destroy scalability.
> 
> > From my purely speculative perspective, it seems unavoidable that message-
> > passing happens at some level in order to keep a shared data structure in sync
> > between a large number of processors. In other words, any access to a shared
> > data structure requires some physical copy no matter what the programming
> > language makes it look like.
> 
> Yes. The hardware maintains the illusion of shared memory so the programmer only need optimize on the basis of an accurate cost model. There is no need to burden the programmer with message passing and manual memory management.

I don't understand this last sentence. Message passing is just an
abstract concept (i.e. an interface) which can e.g. be implemented with
a shared buffer (or the illusion thereof) or other data structure that
decouples sender and receiver processes. Are you against using the right
interfaces when they are appropriate? We don't want to burden the
programmer with too low-level details, do we?

Gerd


> 
> Cheers,
> Jon.
> 
> 
> 
> 


-- 
------------------------------------------------------------
Gerd Stolpmann, Bad Nauheimer Str.3, 64289 Darmstadt,Germany 
gerd@gerd-stolpmann.de          http://www.gerd-stolpmann.de
Phone: +49-6151-153855                  Fax: +49-6151-997714
------------------------------------------------------------


  parent reply	other threads:[~2011-04-20 20:31 UTC|newest]

Thread overview: 74+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <2054357367.219171.1300974318806.JavaMail.root@zmbs4.inria.fr>
2011-03-24 23:13 ` Fabrice Le Fessant
2011-03-25  0:23   ` [Caml-list] " Sylvain Le Gall
2011-03-25  9:55   ` [Caml-list] " Alain Frisch
2011-03-25 11:44     ` Gerd Stolpmann
     [not found]   ` <1396338209.232813.1301046980856.JavaMail.root@zmbs4.inria.fr>
2011-03-25 10:23     ` Fabrice Le Fessant
2011-03-25 12:07       ` Gerd Stolpmann
2011-04-16 12:12         ` Jon Harrop
2011-03-25 10:51   ` Hugo Ferreira
2011-03-25 12:25     ` Gerd Stolpmann
2011-03-25 12:58       ` Hugo Ferreira
     [not found]       ` <341494683.237537.1301057887481.JavaMail.root@zmbs4.inria.fr>
2011-03-25 13:10         ` Fabrice Le Fessant
2011-03-25 13:41           ` Dario Teixeira
2011-03-30 18:12             ` Jon Harrop
2011-03-25 15:44           ` Hugo Ferreira
2011-03-25 18:24             ` Martin Jambon
2011-03-25 19:19               ` Hugo Ferreira
2011-03-25 20:26                 ` Gerd Stolpmann
2011-03-26  9:11                   ` Hugo Ferreira
2011-03-26 10:31                   ` Richard W.M. Jones
2011-03-30 16:56                     ` Jon Harrop
2011-03-30 19:24                       ` Richard W.M. Jones
2011-04-20 21:44                   ` Jon Harrop
2011-04-19  9:57                 ` Eray Ozkural
2011-04-19 10:05                   ` Hugo Ferreira
2011-04-19 20:26                   ` Gerd Stolpmann
2011-04-20  7:59                     ` Hugo Ferreira
2011-04-20 12:30                       ` Markus Mottl
2011-04-20 12:53                         ` Hugo Ferreira
2011-04-20 13:22                           ` Markus Mottl
2011-04-20 14:00                       ` Edgar Friendly
2011-04-19 22:49                 ` Jon Harrop
2011-03-30 17:02               ` Jon Harrop
2011-04-20 19:23               ` Jon Harrop
2011-04-20 20:05                 ` Alexy Khrabrov
2011-04-20 23:00                   ` Jon Harrop
     [not found]                   ` <76544177.594058.1303341821437.JavaMail.root@zmbs4.inria.fr>
2011-04-21  7:48                     ` Fabrice Le Fessant
2011-04-21  8:35                       ` Hugo Ferreira
2011-04-23 17:32                         ` Jon Harrop
2011-04-21  9:09                       ` Alain Frisch
     [not found]                         ` <20110421.210304.1267840107736400776.Christophe.Troestler+ocaml@umons.ac.be>
2011-04-21 19:53                           ` Hezekiah M. Carty
2011-04-22  8:34                           ` Alain Frisch
     [not found]                         ` <799994864.610698.1303412613509.JavaMail.root@zmbs4.inria.fr>
2011-04-22  8:06                           ` Fabrice Le Fessant
2011-04-22  9:11                             ` Gerd Stolpmann
2011-04-23 10:17                               ` Eray Ozkural
2011-04-23 13:47                                 ` Alexy Khrabrov
2011-04-23 17:39                                   ` Eray Ozkural
2011-04-23 20:18                                     ` Alexy Khrabrov
2011-04-23 21:18                                       ` Jon Harrop
2011-04-24  0:33                                       ` Eray Ozkural
2011-04-28 14:42                                         ` orbitz
2011-04-23 19:02                                 ` Jon Harrop
2011-04-22  9:44                             ` Vincent Aravantinos
2011-04-21 10:09                       ` Philippe Strauss
2011-04-23 17:44                         ` Jon Harrop
2011-04-23 17:05                       ` Jon Harrop
2011-04-20 20:30                 ` Gerd Stolpmann [this message]
2011-04-20 23:33                   ` Jon Harrop
2011-03-25 20:27             ` Philippe Strauss
2011-04-19 22:47           ` Jon Harrop
     [not found]           ` <869445701.579183.1303253283515.JavaMail.root@zmbs4.inria.fr>
2011-04-20  9:25             ` Fabrice Le Fessant
2011-03-25 18:45   ` Andrei Formiga
2011-03-30 17:00     ` Jon Harrop
2011-04-13  3:36   ` Lucas Dixon
2011-04-13 13:01     ` Gerd Stolpmann
2011-04-13 13:09       ` Lucas Dixon
2011-04-13 23:04       ` Goswin von Brederlow
2011-04-16 13:54     ` Jon Harrop
2011-03-24 13:44 Alexy Khrabrov
2011-03-24 14:57 ` Gerd Stolpmann
2011-03-24 15:03   ` Joel Reymont
2011-03-24 15:28     ` Guillaume Yziquel
2011-03-24 15:48       ` Gerd Stolpmann
2011-03-24 15:38     ` Gerd Stolpmann
2011-03-25 19:49   ` Richard W.M. Jones

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1303331453.8429.1282.camel@thinkpad \
    --to=info@gerd-stolpmann.de \
    --cc=caml-list@inria.fr \
    --cc=jonathandeanharrop@googlemail.com \
    --cc=martin.jambon@ens-lyon.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).