Mersenne Digest         Monday, June 19 2000         Volume 01 : Number 749




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

Date: Sat, 17 Jun 2000 17:54:08 EDT
From: "Nathan Russell" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Factoring Assignments: Are They Always "First-Time?"

>From: Jeff Woods <[EMAIL PROTECTED]>
>To: [EMAIL PROTECTED]
>Subject: Re: Mersenne: Factoring Assignments:  Are They Always  
>"First-Time?"
>Date: Sat, 17 Jun 2000 17:14:00 -0400
>
>At 01:00 PM 6/17/00 -0700, you wrote:

(snip)

>>If a factor is
>>found for an exponent, it's eliminated from further testing
>>of any kind.
>
>Isn't the factor itself verified?

I would assume it is, however verifying a factor takes well under a P-90 
second.

Nathan
________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com

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

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

Date: Sat, 17 Jun 2000 23:19:15 +0200
From: Lem Novantotto <[EMAIL PROTECTED]>
Subject: Re: Mersenne: LOST NUMBERS

On Sat, 17 Jun 2000 17:28:52 +0200 (CEST), in Mersenne_mailing_list
you wrote:
>On Sat, 17 Jun 2000, Lem Novantotto wrote:
>> Hi!
>> I think it would be nice to post the mersenne numbers you're still
>> working to, and that have been reassigned, so, hopefully, the new
>> tester can read them and stop his testing. Otherwise HE is going to
>> loose time and credit.
>Since first and double checks are credited equally(as far as I know),
>both should get credited with the same amount of work except for counting 
>tests.

Yes. But I'm saying is that, in my experience, if you send a
LL_first_checking result after someone else, your test isn't considered
a double checking... astonishing enough, it isn't considered at all.
Otherwise, how do you explain what happened to me? See my post in reply
to:
Mersenne: assignment 'stolen'
a few posts ago.
- -- 
Bye.
      Lem
- ---------- 'CLOCK is what you make of it' ----------
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Sat, 17 Jun 2000 15:23:04 -0700
From: Spike Jones <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Pointers on farming

Michael Bell wrote:

> ...back to your 37pence/MHz figure.  Also note memory prices have gone up, so
> 32M now costs about �40-50!!  Maybe this is the first time in recent history
> the price of a computer has stabilised??

Hmm, dont think so.  I was over at Fry's electronics this morning.  The 400 MHz
machines like the one I paid 2400 bucks for less than 2 yrs ago with the same
amount of memory was selling for 300 bucks.   Which I think is about 200 pounds?

The 2000-2300 dollar machines are all 1 Ghz now.  Computer prices look to
me to be as much in freefall as ever.  spike

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

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

Date: Sat, 17 Jun 2000 20:14:04 +0100
From: "Siegmar Szlavik" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: LOST NUMBERS

On Sat, 17 Jun 2000 17:28:52 +0200 (CEST), Henrik Olsen wrote:
>On Sat, 17 Jun 2000, Lem Novantotto wrote:
>> On Sat, 17 Jun 2000 05:37:06 -0400, in Mersenne_mailing_list you
>> wrote:
>> >Is all your hard work(and subsequent credit) lost? No, although the server 
>> >has already re-assigned them to another account it will still accept the 
>> >results and properly credit your account for the completed work. 
>> Hi!
>> I think it would be nice to post the mersenne numbers you're still
>> working to, and that have been reassigned, so, hopefully, the new
>> tester can read them and stop his testing. Otherwise HE is going to
>> loose time and credit.
>Since first and double checks are credited equally(as far as I know),
>both should get credited with the same amount of work except for counting 
>tests.
>Time spent on a doublecheck is only wasted if it's turned into a
>triplecheck and all three results agree.
>
Yes, but they are testing in the 10-million-digit range, so it is not
only a question of time and credit. I think it would be just fair to
inform the new testers that they are doing in fact just a doublecheck
and let them decide if they want to continue under these circumstances
or take another 'fresh' exponent.

Siegmar


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

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

Date: Sun, 18 Jun 2000 05:40:42 +0700 (ICT)
From: Warut Roonguthai <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Pointers on farming

On Sat, 17 Jun 2000, Michael Bell wrote:

> Warning on K6's:  As far as GIMPS is concerned they're not too good, because
> the FPU is about half the speed of the Intel Pentium FPU.  (If you run RC5
> then they're as good, if not better, because they have very good integer
> units).

K6 is not good at RC5 either; see http://www.pcbenchmarks.com/distribu.htm
I've heard that RC5 requires some kind of rotate function that is
hardwired on Intel processors but not on K6 and Alpha.

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

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

Date: Sat, 17 Jun 2000 16:09:32 -0700
From: Eric Hahn <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Factoring Assignments:  Are They Always  "First-Time?"

Jeff Woods wrote:
>>being found.  Currently, all exponents thru Prime95's limit of
>>79.3M have been factored to at least 2^50...  If a factor is
>>found for an exponent, it's eliminated from further testing
>>of any kind.
>
>Isn't the factor itself verified?

Yes, it is.  However, at least in the case of Prime95, George
has written the code such that the factor is validated before
it's even displayed as a being a factor and written to the
results file.  If it's invalid, the code continues as if
the "factor" was never found...

Eric




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

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

Date: Sat, 17 Jun 2000 21:56:09 EDT
From: [EMAIL PROTECTED]
Subject: Mersenne: Re: Mersenne Digest V1 #748

<<Willamette(s) which is supposed to debut at 1.4GHz...>>

Remind me what a Willamette is again.  All I know about are Merced (I mean, 
Itanium), and the second-generation Itanium called McKinley.

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

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

Date: Sat, 17 Jun 2000 22:50:04 EDT
From: [EMAIL PROTECTED]
Subject: Mersenne: Prime64?

Will Prime95 be rewritten to run on the Itanium, when it comes out?  Seems to 
 me like 64-bit operation will speed it up significantly, as will the insane 
 amount of registers and floating point units and all the other 
microprocessor 
 whatnot that I'm not current on.  A review (thanks, Stefan Struiker!) of the 
 Itanium mentions, "The CPU will then switch to 32-bit mode on the fly and 
 carry on as if it were a more powerful PIII or Willamette. This all happens 
 because the Itanium supports the IA-32 instructions natively.... All 
software 
 has to be rewritten to take advantage of the IA-64 architecture."  So 
Prime95 
 will run on an Itanium, but a "Prime64" might be even better and wickedly 
 faster.  With what I'm learning about programming, it seems that more 
 registers are always a good thing, especially for memory-intensive processes.
 
 STL
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Sat, 17 Jun 2000 20:17:22 -0700
From: "John R Pierce" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Mersenne Digest V1 #748

Williamette is the next-generation x86 architecture machine.  It has (will
have) a way accelerated CPU clock and even more instructions execute in
fewer clocks, but it has a deeper pipeline and increased instruction
latency.  First generation Williamettes are supposedly going to debut around
1.2GHz and go up from there.  It also has more and better onchip cache, both
L1 and L2.  I believe it has some major structural changes to the PPro/P2/P3
style bus, requiring completely new chipsets.

- -jrp



- ----- Original Message -----
From: <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Saturday, June 17, 2000 6:56 PM
Subject: Mersenne: Re: Mersenne Digest V1 #748


> <<Willamette(s) which is supposed to debut at 1.4GHz...>>
>
> Remind me what a Willamette is again.  All I know about are Merced (I
mean,
> Itanium), and the second-generation Itanium called McKinley.
>
> STL
> _________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

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

Date: Sat, 17 Jun 2000 20:27:01 -0700
From: Stefan Struiker <[EMAIL PROTECTED]>
Subject: Mersenne: Linux And The Slippery Gnome

TeamG:

Trying to start a PrimeHunt on a Linux box, but can't find the Gnome
switch/requeste(o)r  to get the screen resolution down enough so I can read without
a microscope.  Can anyone help?   Am running Redhat 6.1 with a VooDoo
3500 GFX card.

So until last year I thought Linux was a cartoon character.
Leave and learn..

Best Wishes,
Stefanovic


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

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

Date: Sun, 18 Jun 2000 01:27:40 -0400 (EDT)
From: Jason Stratos Papadopoulos <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Prime64?

On Sat, 17 Jun 2000 [EMAIL PROTECTED] wrote:

> Will Prime95 be rewritten to run on the Itanium, when it comes out?  Seems to 
>  me like 64-bit operation will speed it up significantly, as will the insane 
>  amount of registers and floating point units and all the other 
> microprocessor 
>  whatnot that I'm not current on.

Intel has had documentation available on the Web for a while that details
the architecture of Itanium. If Prime95 is ever rewritten for IA64 it
would have to be not only a total rewrite but a complete rethinking of
the FFTs it uses. For example, integer multiply-adds take only a little
longer than floating point multiply-adds; should IA64Prime use an integer
or floating point FFT? If integer, there are big delays in shuffling
between integer and FPU registers (only the FPU can multiply). If floating
point, loads and stores will all take longer, the cache behavior is 
totally different, and the arrays involved get longer because you can't 
pack bits as densely as an integer solution.

Itanium can do two FPU operations per clock, but both can be multiply-adds
instead of just multiplies or adds. Can you rearrange a real-valued FFT
to use multiply-adds as much as possible? It could cut the operation count
in half if you do, but to my knowledge no one has yet done so. It's been
done for complex FFTs, but Prime95 really wants real-valued FFTs, not
complex ones.

Moving to IA64 will be a much bigger challenge than simply rewriting
half a meg of assembly language.

jasonp


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

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

Date: Sun, 18 Jun 2000 04:39:47 -0700
From: Paul Leyland <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Factoring with ECM

> From: Yann Forget [mailto:[EMAIL PROTECTED]]

> Could you explain this to me ?

I can try.  The text below is deliberately informal in style and far from
mathematically rigorous.  The mathematicians on the list will have to
forgive me; those who want a more rigorous explanation should be able to
find one with relative ease.

> I am factoring Fermat numbers with ECM.
> Is this relevant in this case ?

It is relevant for all numbers, but only when ECM produces a composite
factor.  To be honest, your chance of finding even one unknown factor of a
Fermat number with ECM is tiny, and your chances of finding two or more *on
the same curve* are somewhere between nil and negligible.

> Paul Leyland said:
> As long as the coefficients of the curve and the starting point are
> recorded, we can re-run exactly the same computation, with the small
primes
> curtailed as in the p-1 case, on the same curve and the number c.  It's
> because my software doesn't normally output the curve and starting point
> used that the idea hadn't occurred to me.

Ok, what's happening with ECM is that we are performing arithmetic not with
ordinary integers but with points on an elliptic curve --- a polynomial of
the form y^2 = ax^3+b.  (Actually, we are calculating with integers, but in
a complicated manner which corresponds to manipulating points on a
polynomial).  The "starting point" I mentioned is a point which lies on the
curve and, with a few special exceptions, it doesn't matter too much which
point is chosen.  When the arithmetic is performed modulo a composite
number, there are a finite number of points on the curve.  Each different
curve will, in general, have a different number of points.  When the number
of points on a particular curve is divisible *only* by primes less than the
first stage limit B1, with perhaps at most one prime larger than B1 but
smaller than the 2nd stage limit B2, ECM will find a corresponding factor.

Usually this factor will be a prime, but occasionally it will be the product
of two or more primes.   Each of these primes corresponds to a different
selection of small primes smaller than B1 and, usually, only one will
correspond to the prime larger than B1 and smaller than B2.   If you then
repeat the calculation, using exactly the same curve and starting point but
with a reduced set of primes <B1 and/or omit the one >B1, the factor which
corresponds to the omitted primes *won't* be discovered, thereby revealing
the other.


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

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

Date: Sun, 18 Jun 2000 11:23:53 +0200
From: Lem Novantotto <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Mersenne Digest V1 #748

On Sat, 17 Jun 2000 21:56:09 EDT, in Mersenne_mailing_list you wrote:
>Remind me what a Willamette is again.  All I know about are Merced (I mean, 
>Itanium), and the second-generation Itanium called McKinley.

Hi!
If you are interested in differences between willamette and mustang,
try this (a bit technical): http://www.chip-architect.com/mw.pdf
- -- 
Bye.
      Lem
- ---------- 'CLOCK is what you make of it' ----------
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Sun, 18 Jun 2000 10:44:41 -0700
From: Bob Margulies <[EMAIL PROTECTED]>
Subject: Mersenne: New DSL connection

The DSL modem on my Windows 98 PC works beautifully. But now when
Prime95 tries to contact the server, I get the message

Dial-up connection not active.
Will try contacting server again in 2 minutes.

What do I need to do?
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Sun, 18 Jun 2000 13:13:56 -0700
From: Will Edgington <[EMAIL PROTECTED]>
Subject: Mersenne: Linux And The Slippery Gnome

   Trying to start a PrimeHunt on a Linux box, but can't find the
   Gnome switch/requeste(o)r to get the screen resolution down enough
   so I can read without a microscope.  Can anyone help?  Am running
   Redhat 6.1 with a VooDoo 3500 GFX card.

Presuming you're running in an xterm window, try holding down a control
key and pressing & holding your third mouse button while in that
window; xterm should bring up a list of font sizes.  Simply release the
mouse button while the mouse cursor is on the one that you want to try.
This means that each xterm can have a different font size, if that's
what you want.  The font, including size, can also be given on the
command line, as in 'xterm -fn 9x15bold' (e.g.).  'xlsfonts' will list
the fonts your X11 server knows about, but it's usually a very long
list, so you probably want to redirect it into a file ('xlsfonts >
file') or something similar ('xlsfonts | more', perhaps).

Setting the font you want so that new xterms start with it is somewhat
messy, usually involving your ~/.Xdefaults file, but there may be a
configuration tool in Gnome that I'm not aware of (I was using X11
before Linux was created, let alone before Gnome, so ... :).

Some X11 configurations, usually declared in /etc/X11/XF86Config,
support more than one video resolution, which is usually changed while
running X with control-alt-(keypad's + key) and control-alt-(keypad -).
These support virtual screens in that the pixels that fit within the
display area are only part of what X11 is actually displaying; simply
try to move the mouse off the edge of the screen towards what you want
to see that's off the edge.  The size of the virtual screen is only
limited by the amount of memory your video card can access and the
color depth (256 colors fit in 8 bits, e.g., as Mersenne hunters
should know :).

                                                        Will

http://www.garlic.com/~wedgingt/mersenne.html
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Sun, 18 Jun 2000 13:58:59 -0700
From: Will Edgington <[EMAIL PROTECTED]>
Subject: Mersenne: Factoring Assignments:  Are They Always "First-Time?"

Stefan Struiker writes:
   When a requested factoring assignment is listed with, say, 52 in an
   account log, does this mean it has been factored to 52 bits, but
   _without_ success?

Yes, the number should have no factors less than 2^52.

   Or could a factor have already been found in some cases, but less
   than 52 bits long?

Nope, unless the factor was not reported for some reason (bug, disk
crash, etc.).

   My strategy in factoring 13.3 mill exponents and up, is to save L-L
   testing and DCing time by knocking some out early.  Seem to be on a
   roll, too, with factors found 40% of the time, with a turnaround of
   40 hours per.

That's a very high rate of factors, I'd've thought, but that happens
sometimes.

In any case, Prime95 "knows" how much factoring work should be done
for a particular Mersenne number before starting an LL test (first or
double-check) on it and will do more factoring if the data it gets
from Primenet (or other source) indicates the number has not been
factored "enough".  The predicated chances of finding a factor during
trial and P-1 factoring is taken into account, along with how long the
factoring takes to do and how long the two LL tests will take.

So your phrase "knocking some out early" is exactly correct: if noone
tries to factor a particular Mersenne number before it is given to a
Prime95 that wants to run an LL test, that Prime95 will do some
factoring first, usually before it even finishes the prior Mersenne
number's LL test (to make sure it has "enough" work in worktodo.ini).

Eric Hahn writes:
   If it's listed as 52 in the fact-bits column of the report, it
   means that it's been trial-factored thru 2^52 without any factors
   being found.  Currently, all exponents thru Prime95's limit of
   79.3M have been factored to at least 2^50...  If a factor is
   found for an exponent, it's eliminated from further testing
   of any kind.

Yup.  Here's a short summary of my current data.  For Mersenne numbers
with prime exponent that have no LL test nor a factor, here are the
smallest exponents trial factored only as far as the last column:

M( 5178743 )U: 2^62
M( 8896813 )U: 2^61
M( 9993539 )U: 2^60
M( 10078559 )U: 2^55
M( 11300657 )U: 2^54
M( 11505331 )U: 2^53
M( 11521879 )U: 2^52
M( 20500019 )U: 2^51
M( 30100181 )U: 2^50
M( 79300037 )U: 2^45
M( 79306169 )U: 2^43

The exponents above 79.3 million have probably only been worked on by
me, personally, since they're above Prime95's limit, but I'm still a
bit surprised they haven't been factored further; trial factoring to
the same depth is _easier_ for larger exponents, not harder.

Jeff Woods writes:
   Isn't the factor itself verified?

Yes, if only by me, as I noted in another thread in the last couple of
weeks.

                                                Will

http://www.garlic.com/~wedgingt/mersenne.html
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Sun, 18 Jun 2000 15:22:01 +0200
From: "Steinar H. Gunderson" <[EMAIL PROTECTED]>
Subject: Mersenne: Re: Pointers on farming

On Sun, Jun 18, 2000 at 05:40:42AM +0700, Warut Roonguthai wrote:
>K6 is not good at RC5 either; see http://www.pcbenchmarks.com/distribu.htm
>I've heard that RC5 requires some kind of rotate function that is
>hardwired on Intel processors but not on K6 and Alpha.

But K6 is quite OK at factoring, since there is a special integer version of
factoring in Prime95.

/* Steinar */
- -- 
Homepage: http://members.xoom.com/sneeze/
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Sun, 18 Jun 2000 22:41:24 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: LOST NUMBERS

On 17 Jun 00, at 23:19, Lem Novantotto wrote:

> >Since first and double checks are credited equally(as far as I know),
> >both should get credited with the same amount of work except for counting 
> >tests.
> 
> Yes. But I'm saying is that, in my experience, if you send a
> LL_first_checking result after someone else, your test isn't considered
> a double checking... astonishing enough, it isn't considered at all.

AFAIK PrimeNet will record the result of the second test - though it 
may not be obvious from the reports - and you won't get the CPU time 
credit for it. The result will work its way into the database (look 
in lucas_v.txt) & will be credited in George's list (which also 
includes manually submitted results).


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

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

Date: Sun, 18 Jun 2000 19:54:07 -0400
From: "Vincent J. Mooney Jr." <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Prime64?

I met George Woltman in Orlando, Florida a week ago and as I recall, he
agreed that (as you said, Jason) "If Prime95 is ever rewritten for IA64 it
would have to be not only a total rewrite but a complete rethinking of the
FFTs it uses."  George is up to the task, happily. 


At 01:27 AM 6/18/00 -0400, Jason Papad wrote:
>
>
>On Sat, 17 Jun 2000 [EMAIL PROTECTED] wrote:
>
>> Will Prime95 be rewritten to run on the Itanium, when it comes out?
Seems to 
>>  me like 64-bit operation will speed it up significantly, as will the insane 
>>  amount of registers and floating point units and all the other 
>> microprocessor 
>>  whatnot that I'm not current on.
>
>Intel has had documentation available on the Web for a while that details
>the architecture of Itanium. If Prime95 is ever rewritten for IA64 it
>would have to be not only a total rewrite but a complete rethinking of
>the FFTs it uses. For example, integer multiply-adds take only a little
>longer than floating point multiply-adds; should IA64Prime use an integer
>or floating point FFT? If integer, there are big delays in shuffling
>between integer and FPU registers (only the FPU can multiply). If floating
>point, loads and stores will all take longer, the cache behavior is 
>totally different, and the arrays involved get longer because you can't 
>pack bits as densely as an integer solution.
>
>Itanium can do two FPU operations per clock, but both can be multiply-adds
>instead of just multiplies or adds. Can you rearrange a real-valued FFT
>to use multiply-adds as much as possible? It could cut the operation count
>in half if you do, but to my knowledge no one has yet done so. It's been
>done for complex FFTs, but Prime95 really wants real-valued FFTs, not
>complex ones.
>
>Moving to IA64 will be a much bigger challenge than simply rewriting
>half a meg of assembly language.
>
>jasonp
>
>
>_________________________________________________________________
>Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
>Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers
>

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

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

Date: Sun, 18 Jun 2000 20:15:33 -0400
From: "Vincent J. Mooney Jr." <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Factoring

Thanks for the factor.exe citation.

At 01:40 AM 6/16/00 -0700, Jim Howell wrote:
>[Wed 14 Jun 2000, Paul Leyland writes]
>
>Today I found this number 3756482676803749223044867243823 with ECM and
>B1=10,000.  It has two factors, each of 16 digits, which could *not* have
>been found by trial division in any reasonable time.
>
>-------------
>
>I use a program called "factor.exe", which uses several factoring methods.
It factors the above number within several seconds.  (For this number, the
factors are found with the P-1 method.)  In case anyone is interested, the
factors are  1483398061194277 and 2532349728015299.
>
>This program runs on Windows, and can be downloaded from Chris Caldwell's
main page, at:
>
>http://www.utm.edu/research/primes
>
>Go down to section 4, (Software), and look for "factor.exe", described as a
DOS program, but it actually runs in a Command Window on Windows 95 and
later, and (probably) not under actual DOS.  I find "factor.exe" quite
useful for factoring small numbers (it will accept numbers up to about 130
digits).
>--Jim
>
><!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
><HTML><HEAD>
><META content="text/html; charset=iso-8859-1" http-equiv=Content-Type>
><META content="MSHTML 5.00.2314.1000" name=GENERATOR>
><STYLE></STYLE>
></HEAD>
><BODY bgColor=#ffffff>
><DIV><FONT face=Arial size=2>[Wed 14 Jun 2000, Paul Leyland
writes]</FONT></DIV>
><DIV><FONT face=Arial size=2><BR>Today I found this number 
>3756482676803749223044867243823 with ECM and<BR>B1=10,000.&nbsp; It has two 
>factors, each of 16 digits, which could *not* have<BR>been found by trial 
>division in any reasonable time.</FONT></DIV>
><DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
><DIV><FONT face=Arial size=2>-------------</FONT></DIV>
><DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
><DIV><FONT face=Arial size=2>I use a program called "factor.exe", which uses 
>several factoring methods.&nbsp; It factors the above number&nbsp;within
several 
>seconds.&nbsp; (For this number, the factors are found with the P-1 
>method.)&nbsp; In case anyone is interested, the factors are&nbsp; 
>1483398061194277 and 2532349728015299.<BR></FONT></DIV>
><DIV><FONT face=Arial size=2>This program runs on Windows, and can be
downloaded 
>from Chris Caldwell's main page, at:</FONT></DIV>
><DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
><DIV><FONT face=Arial size=2><A 
>href="http://www.utm.edu/research/primes">http://www.utm.edu/research/prime
s</A></FONT></DIV>
><DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
><DIV><FONT face=Arial size=2>Go down to section 4, (Software), and look for 
>"factor.exe", described as a DOS program, but it actually runs in a Command 
>Window on Windows 95 and later, and (probably) not under actual DOS.&nbsp; I 
>find "factor.exe" quite useful for factoring small numbers (it will accept 
>numbers up to about 130 digits).</FONT></DIV>
><DIV><FONT face=Arial size=2>--Jim</FONT></DIV>
><DIV><FONT face=Arial size=2>&nbsp;</DIV></FONT></BODY></HTML>
>

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

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

Date: Sun, 18 Jun 2000 17:40:09 -0700
From: "Pardoe, Richard (PRDR)" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: New DSL connection

Open the Prime95 window
Select Test - PrimeNet

Remove the check mark before "Use a dial-up connection to the Internet"

From: Bob Margulies
>The DSL modem on my Windows 98 PC works beautifully. But now when
>Prime95 tries to contact the server, I get the message
>
>Dial-up connection not active.
>Will try contacting server again in 2 minutes.
>
>What do I need to do?


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

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

Date: Sun, 18 Jun 2000 21:25:11 -0400
From: "Larry Murray" <[EMAIL PROTECTED]>
Subject: Mersenne: LOST NUMBERS

I thought had occurred to me with this prize that is being offered.  what happens if 
you work on a number for 6 months and then it is re-assigned to a person with a faster 
computer and their computer finishes the computation first and it is 
found to be a prime.  Who is entitled to the Prize? Does that mean if you have A 
slower 550mhz computer don't bother testing a 10 million number? If you do make sure 
you always check in because it can be taken from you?  On the same subject what-if 
you're the person the number is reassigned to and you work on a number for 6
 months  to find that it was a reassignment and it was found to be prime...I think 
with the longer testing time of the 10 million digit numbers the time between 
reassignment of those numbers should be much longer.  I personally have been running 3 
ten million digit numbers on Pentium 550's since September of 1999 and hardly 
ever even bother with the computers that are running them I have strictly devoted them 
to the 
mersenne project...ANYWAY ITS SOMETHING TO THINK ABOUT--TAKE CARE ALL
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ      -- http://www.tasam.com/~lrwiman/FAQ-mers

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

Date: Mon, 19 Jun 2000 13:21:13 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: LOST NUMBERS

On 18 Jun 00, at 21:25, Larry Murray wrote:

> I thought had occurred to me with this prize that is being offered.  what
> happens if you work on a number for 6 months and then it is re-assigned to
> a person with a faster computer and their computer finishes the
> computation first and it is found to be a prime.  Who is entitled to the
> Prize? Does that mean if you have A slower 550mhz computer don't bother
> testing a 10 million number? If you do make sure you always check in
> because it can be taken from you?  On the same subject what-if you're the
> person the number is reassigned to and you work on a number for 6
>  months  to find that it was a reassignment and it was found to be
>  prime...I think with the longer testing time of the 10 million digit
>  numbers the time between reassignment of those numbers should be much
>  longer.  I personally have been running 3 ten million digit numbers on
>  Pentium 550's since September of 1999 and hardly 
> ever even bother with the computers that are running them I have strictly
> devoted them to the mersenne project...ANYWAY ITS SOMETHING TO THINK
> ABOUT--TAKE CARE ALL

You have to check in occasionally to keep the assignment. This is 
entirely reasonable since some people are bound to "default" without 
bothering to return the assignment.

If it's too inconvenient to connect the actual systems to allow them 
to check in, you can use the PrimeNet Manual Testing page to check 
assignments in manually in order to prevent them from expiring. Since 
you can "extend" an assignment for up to 120 days (plus the 60 day 
grace period) you should only need to do this two or three times 
during the run.

If you run assignments which aren't given to you by PrimeNet, or 
continue to work on assignments which have been reallocated due to 
your failure to check them in occasionally, I don't think you should 
be entitled to a share of any prize. However it looks as though, 
according to the EFF rules (which I haven't looked at for a while), 
the first discovery reported to EFF takes precedence.

Since the expected return on testing numbers in the 10 million digit 
range is of the order of 40 cents / PIII-550 year, I doubt too many 
people are participating simply because of the existence of the prize 
... ?

Also, note that it's entirely possible that the EFF prize will be won 
by someone working on non-Mersenne numbers using entirely different 
software and/or hardware.

BTW for QA reasons I am already working on a double-check of a 10 
million digit number before the first test is completed. I will make 
damn sure that the "official" owner of the assignment reports the 
final result, abandons or allows the assignment to expire before I 
report the result myself. (And the "official" owner knew this before 
I started!)

Personally, and bearing in mind that there are a lot of much smaller 
exponents which still require testing, I consider a standard PC 
running a PIII-550 to be inadequate for running 10 million digit 
exponents. I'm using an Athlon 650, which is about 35% faster, and I 
have a system (self-)built for reliability rather than down to a 
price. However, GOOD LUCK to all those who do chance their arm!

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

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

Date: Mon, 19 Jun 2000 13:21:13 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Prime64?

On 18 Jun 00, at 1:27, Jason Stratos Papadopoulos wrote:

> For example, integer multiply-adds take only a little longer
> than floating point multiply-adds; should IA64Prime use an integer or
> floating point FFT? If integer, there are big delays in shuffling between
> integer and FPU registers (only the FPU can multiply). If floating point,
> loads and stores will all take longer, the cache behavior is totally
> different, and the arrays involved get longer because you can't pack bits
> as densely as an integer solution.

This obviously needs to be looked at & evaluated on real hardware.
> 
> Itanium can do two FPU operations per clock, but both can be multiply-adds
> instead of just multiplies or adds. Can you rearrange a real-valued FFT to
> use multiply-adds as much as possible? It could cut the operation count in
> half if you do, but to my knowledge no one has yet done so.

Yes, but you get only about 40% saving - you need about 10 additions 
for every 6 multiplications. "No-one has bothered" because the effort 
is pointless on IA32 architecture processors, and there simply aren't 
enough processors which have an efficient "a*b+c" instruction around 
to make it worthwhile to put in the required programming effort.

(What goes around, comes around. The VAX architecture - very, very 
CISC - had three-operand instructions, but the trend to RISC systems 
made such things unfashionable. Now hardware designers are realizing 
that we need such things for performance reasons!)

> Moving to IA64 will be a much bigger challenge than simply rewriting
> half a meg of assembly language.

I wonder whether we will get better performance running IA32 software 
or using a HLL program like Mlucas in native IA64 mode? Depends to a 
large extent on the availability & performance of optimizing 
compilers, I suppose.

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

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

Date: Mon, 19 Jun 2000 20:43:11 +0200
From: Guillermo Ballester Valor <[EMAIL PROTECTED]>
Subject: Mersenne: New Lucas-Lehmer test program.

Hi to all:

In the last weeks I've been writing a program to perform l-l test. I'm
the guy who wrote an adaptation of MacLucasUNIX to FFTW package half
year ago. It seemed to run fast on intelx86  but in other platforms the
Mlucas and MacLucasUNIX was better than MacLucasFFTW.

I had learnt from that the first thing I must to made was to write my
own FFT code. I did it, and the code seems to be fast. I called it
YEAFFT (Yet Another FFT) and actually is a fast convolver. It is a
complex based FFT and uses intensively the C-macro expansion facilities.
The FFT routines are based in a short file of macros which includes all
the common tasks. This macros now are written in a generic form but I
suppose they can be tuned for a particular target, even more it can be
replaced by assembler code if our C-compiler supports this feature. 

My aim when writing the code was that it could run I a the big variety
of systems. In fact, my test version (without priority management) runs
in a old 486 pc with Borland turbo c++ under MSdos 6-20, and also runs
on Alpha 21264 . With the -Dmacro compiler flags one can choose a lot of
features of the package to adjust it to the target system and made it as
fast as possible.   
 
Ernst W. Mayer kindly has made some timings and has created a telnet
count for me in his machines. Without his suggestions and help in
testing I've not been able to write this code. The results are similar
to Mlucas on MIPS and Alpha's. It can be a good new for those platforms
without a f90 compiler but good C ones (for example Apple Mac. users).
It supports the same FFT lengths than prime95 plus a radix-9 (like
Mlucas). The Lucas-Lehmer test code, which calls to YEAFFT routines is
called Glucas. The code runs fine on my pentium 166-MMX under SuSE Linux
(about 50% of performance than mprime, but better than MacLucasFFTW).
The I/O code is from Will Edgington MERS package. Now Will Edgington is
working to include it in his MERS release, and hope soon we will have
results.

For test proposes there is other program, ylucas2, with the same
interface than original Dr. Crandall lucdwt.c (it has not restart
features nor priority management and then not recommended for a complete
big l-l test) but is good to see the possibilities of Glucas. I can send
a zip copy of the source to interested people. Send me a private e-mail. 

Soon there will be other L-L tester in the GIMPS arena. I wonder whether
there will be a 'universal' interface to primenet, (perhaps in Java),
able to read the outputs from Mlucas, MacLcasUNIX, Glucas ...  dialog
with primenet server and manage the tasks to do by with the clients, all
in automatic form (no manual).    
      
Regards.

Guillermo.
_________________________________________________________________
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 #749
******************************

Reply via email to