Mersenne Digest        Wednesday, April 4 2001        Volume 01 : Number 836




----------------------------------------------------------------------

Date: Mon, 02 Apr 2001 13:45:34 -0500
From: Herb Savage <[EMAIL PROTECTED]>
Subject: Re: Mersenne: reconfigurable MP processor

[EMAIL PROTECTED] wrote:
> 
> This is interesting:
> 
> http://www.cnn.com/2001/TECH/ptech/03/30/langley.supercomputer/index.html
> 
> I wonder what the power dissipation for such a beast is?

250 watts according to their site:

  http://www.starbridgesystems.com/prod-hal1.html

Regards,

Herb Savage
_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

------------------------------

Date: Mon, 2 Apr 2001 22:25:31 +0200
From: "Bjoern Hoffmann" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: P III 750 -E (B) Problem with "MS Me"

Brian J. Beesley wrote:

> > I have a little problem with a PIII 750 MHZ notebook, that
> is running
> > prime on a 11 mio digit. It's really too slow, about a 0.4x per
> > iteration, should be a 0.2x... Where is the problem?
>
> (1) as other people have suggested, powersaving mode. If this is the
> correct assumption, if you keep moving the mouse pointer (once a
> minute should be enough), the system will run at a "sensible" speed.

powersaving is disabled, mouse moving changes nothing.

> (2) something else stealing CPU cycles ... ?

prime95 is working on nearly 100% CPU time.

> (3) are these not the processors with "SpeedStep" i.e. they run at
> about half speed when on battery power?

intel speed step is disabled.

> (4) many notebooks have some sort of system for manually selecting
> the speed from the keyboard. e.g. on a Tosh notebook, try <Fn>+<F2>.
> This cycles through 3 modes, "slow", "fast" & "user", you can set up
> the powersaving CPU speed & screen brightness together with the
> timeout values in user mode using BIOS Setup or a special utility
> supplied with the system.

I can't find any of these settings, but I asked the manufactor of the
notebook.

In the end, the problem remains, the PIII750 is too slow, a 0.4x where a
0.2x should be. Interesting: CPU Benchmark with Dr. Hardware 2000
(Sofstone and Hardstone) is allright, in softstone it is faster, in
hardstone a little bit slower than a Athlon 750.

Maybe there is a problem with Microsofts ME (any of these drivers),
which was installed on the notebook? On nearly every startup there is a
ILLEGAL SUMOUT, but only on startups. There is no problem with it even
after a couple of days running. Has anyone heard of problems with MS Me?
Next week or so I will try Win2k or Linux.

best reagrds from Germany
Bjoern

_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

------------------------------

Date: Mon, 02 Apr 2001 23:59:54 +0200
From: Henk Stokhorst <[EMAIL PROTECTED]>
Subject: Mersenne: graphical image of factoring work being done

L.S.,

I wrote a small program in Delphi that uses the 'nofactor' files 
produced by 'decomp' as input to create a graphical image of the work 
that has been performed on factoring. It creates a scalable image such 
as can be found at

http://home.planet.nl/~tha/overview20010402.gif

The .exe file for windows can be found at

http://home.planet.nl/~tha/overview.zip

it includes the Delphi 5 source files, but you can throw them away if 
you don't find any use for them. To monitor the changes over a period of 
time, say a week I wrote another program which highlights the changes 
made between a first and second generated file, it can be found at

http://home.planet.nl/~tha/changes_overview.zip

For the first program create an input file using the command 'decomp -n 
0 80000000' and then run overview in windows.
For the second program rename the input file from 'nofactor' to 
'nofactor_old', create a new nofactor file with the same range and then 
run factor_changes_overview.

Mail me for any questions.

YotN,

Henk Stokhorst

_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

------------------------------

Date: Tue, 3 Apr 2001 03:59:51 +0200
From: "Hoogendoorn, Sander" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: graphical image of factoring work being done

Henk Stokhorst wrote:

> I wrote a small program in Delphi that uses the 'nofactor' 
> files produced by 'decomp' as input to create a graphical 
> image of the work that has been performed on factoring.

Thanks, this is something i've been looking for, much easier
then importing the file into Excel (which can handle only
65536 lines) and sorting it.

I found a bug in the program though, when you use a nofactor
file which doesn't start with 1 the horizontal scale is
incorrect.

Further, is it possible to stretch the scale so that only the
part for which there is data in de nofactor file is displayed?
And can you tell the difference between the red and black bars?

Thanks

Sander
_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

------------------------------

Date: Tue, 03 Apr 2001 17:36:23 +0200
From: Henk Stokhorst <[EMAIL PROTECTED]>
Subject: Re: Mersenne: graphical image of factoring work being done

Hoogendoorn, Sander wrote:

> I found a bug in the program though, when you use a nofactor
> file which doesn't start with 1 the horizontal scale is
> incorrect.
> 
> Further, is it possible to stretch the scale so that only the
> part for which there is data in de nofactor file is displayed?
> And can you tell the difference between the red and black bars?

Yep, there was a bug in it, and yes, it can be streched. I rewrote the 
application from scratch and it works much better now. You can zoom to 
detail by feeding the program only a minor part of the nofactor data. 
The bug is gone too.

Every dot is red with a black line around it. Many dots adjacent put the 
black over the red, hence a red dot means it is more 'lonely' then the 
black dots.

If you downloaded http://home.wxs.nl/~tha/overview.zip please download 
the new version because of accuracy and speed. The other application has 
not been updated yet.

YotN,

Henk Stokhorst

_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

------------------------------

Date: Wed, 04 Apr 2001 21:06:30 -0700
From: "Gary Untermeyer" <[EMAIL PROTECTED]>
Subject: Mersenne: arithmetic progression of consecutive primes

Greetings,

Although this is not a question regarding mersenne primes, I thought I'd
throw this out to the readers here.

Let x be a prime number.  Consider the series of  numbers that take the
following form:

x,  x + n,  x + 2n,  x + 3n,  x + 4n,  x + 5n,  x + 6n,  where n is an
even positive whole number.

In this series of seven numbers, can anyone tell me why, if ALL of these
numbers are prime, that the minimum value of n is 210 if all the terms
are _consecutive_ prime numbers?

This is not a trick question.  I think it may have something to do with
210 being the product of the first four prime numbers.  After all, the
first instance of six consecutive primes in an arithmetic progression
occurs with a value of 30, which is the product of the first three
primes.  The first instance of four terms occurs with a value of 6.
And, if you can answer that question, how does the answer generalize to
consecutive primes in arithmetic progression when looking for longer
series?  After a gap of 210 for seven terms, is the next gap, for eight
terms, 2310?

I found this stated on the web at:
http://felix.unife.it/Root/d-Mathematics/d-Number-theory/t-7-primes-in-progression

Regards,

Gary Untermeyer

_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

------------------------------

End of Mersenne Digest V1 #836
******************************

Reply via email to