Prev: Inscribed triangles
Next: JSH: Morphic non-linear Factoring Random solved
From: Pubkeybreaker on 12 May 2010 08:47
On May 10, 4:51 am, Mok-Kong Shen <mok-kong.s...(a)t-online.de> wrote:
> Noob wrote:
> > Gerry Myerson wrote:
>
> >> Pubkeybreaker wrote:
>
> >>> Join NFS(a)Home! Help finish the Cunningham project, the longest
> >>> ongoing computational project in history.
>
> >> When you want people to join something, it's traditional to tell them
> >> how.
>
> > The project description on Wikipedia
> >http://en.wikipedia.org/wiki/NFS(a)Home
>
> > The official website
> >http://escatter11.fullerton.edu/nfs/
>
> As I remarked in sci.crypt in a thread (initiated 05.04) of the same
> title, there are practical issues (no software to tightly control
> the CPU consumption and necessity of constant connection with the
> server) that likely renders a participation unattractive for many
> people with computers running OS like Windows.
>

Your reply assumes facts that are contrary to the available evidence.
BOINC supports both Windows and off-line processing. Work
Assignments
are/can be small, so that code only runs for an hour or two, thus
limiting
CPU usage.
From: Tom St Denis on 12 May 2010 08:47
On May 12, 8:30 am, Pubkeybreaker <pubkeybrea...(a)aol.com> wrote:
> On Apr 6, 5:58 am, Mok-Kong Shen <mok-kong.s...(a)t-online.de> wrote:
>
>
>
>
>
> > Pubkeybreaker wrote:
> > > Check outNFS(a)Home;   help them push towards larger numbers.
>
> > >http://escatter11.fullerton.edu/nfs/
>
> > I wouldn't refute being called a selfish person, but I think
> > I would like to take part in some internet collective scientific
> > computing projects, if the following conditions could be satisfied
> > (I don't know much about the actual working conditions of such
> > projects, so part below may be irrelevant):
>
> > 1. One can dynamically set an upper limit of the CPU load of the
> > process.
>
> > 2. One can download on one's initiative the task to be done. During
> > the actual processing there is no need of an internet connection to
> > the server of the project. One need not have one's computer on 24 hours
> > a day, i.e. the process can be interrupted and resumed at any time. One
> > uploads the result oneself, when the task processing comes to an end.
>
> > M. K. Shen
>
> BOINC  already allows you to do this.  The current NFS(a)Home project
> supports
> the features you ask for.

I'd help except my main computer at home is a laptop. I do have a
server [quad core] but really the thought of running it full bore is
not that appealing.

BOINC style projects are for people who don't pay their electricity
bills.

Tom
From: Chip Eastham on 12 May 2010 08:56
On May 12, 8:47 am, Tom St Denis <t...(a)iahu.ca> wrote:
> On May 12, 8:30 am, Pubkeybreaker <pubkeybrea...(a)aol.com> wrote:
>
>
>
>
>
> > On Apr 6, 5:58 am, Mok-Kong Shen <mok-kong.s...(a)t-online.de> wrote:
>
> > > Pubkeybreaker wrote:
> > > > Check outNFS(a)Home;   help them push towards larger numbers.
>
> > > >http://escatter11.fullerton.edu/nfs/
>
> > > I wouldn't refute being called a selfish person, but I think
> > > I would like to take part in some internet collective scientific
> > > computing projects, if the following conditions could be satisfied
> > > (I don't know much about the actual working conditions of such
> > > projects, so part below may be irrelevant):
>
> > > 1. One can dynamically set an upper limit of the CPU load of the
> > > process.
>
> > > 2. One can download on one's initiative the task to be done. During
> > > the actual processing there is no need of an internet connection to
> > > the server of the project. One need not have one's computer on 24 hours
> > > a day, i.e. the process can be interrupted and resumed at any time. One
> > > uploads the result oneself, when the task processing comes to an end.
>
> > > M. K. Shen
>
> > BOINC  already allows you to do this.  The current NFS(a)Home project
> > supports
> > the features you ask for.
>
> I'd help except my main computer at home is a laptop.  I do have a
> server [quad core] but really the thought of running it full bore is
> not that appealing.
>
> BOINC style projects are for people who don't pay their electricity
> bills.
>
> Tom

BOINC seems to use multiple cores/CPUs to run parallel
tasks. On my dual CPU it executes two tasks, but this
is optionally limitable, so that if you want, BOINC can
be restricted to one task & thus one CPU at a time.

--c
From: Tom St Denis on 12 May 2010 09:06
On May 12, 8:56 am, Chip Eastham <hardm...(a)gmail.com> wrote:
> On May 12, 8:47 am, Tom St Denis <t...(a)iahu.ca> wrote:
>
>
>
>
>
> > On May 12, 8:30 am, Pubkeybreaker <pubkeybrea...(a)aol.com> wrote:
>
> > > On Apr 6, 5:58 am, Mok-Kong Shen <mok-kong.s...(a)t-online.de> wrote:
>
> > > > Pubkeybreaker wrote:
> > > > > Check outNFS(a)Home;   help them push towards larger numbers.
>
> > > > >http://escatter11.fullerton.edu/nfs/
>
> > > > I wouldn't refute being called a selfish person, but I think
> > > > I would like to take part in some internet collective scientific
> > > > computing projects, if the following conditions could be satisfied
> > > > (I don't know much about the actual working conditions of such
> > > > projects, so part below may be irrelevant):
>
> > > > 1. One can dynamically set an upper limit of the CPU load of the
> > > > process.
>
> > > > 2. One can download on one's initiative the task to be done. During
> > > > the actual processing there is no need of an internet connection to
> > > > the server of the project. One need not have one's computer on 24 hours
> > > > a day, i.e. the process can be interrupted and resumed at any time. One
> > > > uploads the result oneself, when the task processing comes to an end.
>
> > > > M. K. Shen
>
> > > BOINC  already allows you to do this.  The current NFS(a)Home project
> > > supports
> > > the features you ask for.
>
> > I'd help except my main computer at home is a laptop.  I do have a
> > server [quad core] but really the thought of running it full bore is
> > not that appealing.
>
> > BOINC style projects are for people who don't pay their electricity
> > bills.
>
> > Tom
>
> BOINC seems to use multiple cores/CPUs to run parallel
> tasks.  On my dual CPU it executes two tasks, but this
> is optionally limitable, so that if you want, BOINC can
> be restricted to one task & thus one CPU at a time.

Point is it heats up the server and costs electricity. These sorts of
tasks are better dealt with machines built for the purpose. Not
random peoples file servers....

Tom
From: Mok-Kong Shen on 12 May 2010 16:46
Pubkeybreaker wrote:
> Mok-Kong Shen wrote:

>> As I remarked in sci.crypt in a thread (initiated 05.04) of the same
>> title, there are practical issues (no software to tightly control
>> the CPU consumption and necessity of constant connection with the
>> server) that likely renders a participation unattractive for many
>> people with computers running OS like Windows.
>>
>
> Your reply assumes facts that are contrary to the available evidence.
> BOINC supports both Windows and off-line processing. Work
> Assignments
> are/can be small, so that code only runs for an hour or two, thus
> limiting
> CPU usage.

I accessed http://escatter11.fullerton.edu/nfs/. While the page is all
English, an exception is the item "Unsere Regeln und Grunds�tze" in
German there. One reads the following:

Die Anwendungen von NFS(a)Home k�nnen auf manchen Computern zu
�berhitzung f�hren. Wenn dies passiert, sollten Sie aufh�ren, mit
dem betreffenden Computer f�r NFS(a)Home zu arbeiten oder ein
Hilfprogramm einsetzen, das die Prozessornutzung beschr�nkt.

Isn't that disconcerting enough for a potential participant?

What I would think that most potential doners of computer time with
machines running Windows would like is: there be a mechanism with which
one can set 'at any time' the cpu-time consumption of the
nfs-application to x-%, where x is any arbitrary value between 0 and
100. Are you sure that such a mechanism exists (to be downloaded from
BOINC or elsewhere)?

As to connection to the server, I don't exactly know, but the follow-up
of Chip Eastham of 06.04.2010 19:31 indicated that there had been at
least some problem in that direction.

M. K. Shen





First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: Inscribed triangles
Next: JSH: Morphic non-linear Factoring Random solved