Mersenne Digest        Wednesday, June 9 1999        Volume 01 : Number 571




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

Date: Tue, 08 Jun 1999 06:34:55 -0500
From: [EMAIL PROTECTED] (Mikus Grinbergs)
Subject: Re: Mersenne: Prime 95 Error Messages/ Misc

Note to the QA testers:

I personally do not overclock, but from time to time I do upgrade
motherboards.  I have seen opinions on this mailing list that the
existing 16-hour self-test might not be enough to bring hardware
deficiencies to light.  Someone suggested running the torture
test - but is that more "rigorous" than the self-test?  And why
should running the torture test for a week be any more definitive
than already running Lucas-Lehmer on an assigned exponent?

My suggestion is that the QA people should "update" the built-in
self-test to make it more likely to provoke hardware deficiencies.
(Also, now that we are in the realm of "large" FFTs, is it still
useful for the self-test to spend so much time on "short" FFTs ?)

mikus


In article <000001beb17a$0e5dfe40$5deabfa8@buster>,
"Ethan Hansen" <[EMAIL PROTECTED]> wrote:
>
> It is also possible to have an overclocked CPU pass the full self test
> suite, but later exhibit problems.  The likely culprit is simple wearout --
> the CPU initially was barely functional at the overclocked speed, but slowed
> enough that it no longer runs.  If this happens, you usually can still run
> the CPU at the rated speed.

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 8 Jun 1999 08:27:32 -0600
From: "Aaron Blosser" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Prime 95 Error Messages/ Misc

> If you've overclocked your system at all (_not_ reccomended, but I
> know it can be successful in some cases) then I suggest you let
> the torture test run for a couple of days before committing yourself
> to doing "real" work. It can happen that an overclocked system
> appears to run fine for "office" applications but causes problems
> with Prime95 because very few other applications use the floating-
> point unit even half as intensively as Prime95 does.
>
> Note that simple overheating can also cause problems, even if your
> system _isn't_ overclocked. Might be an idea to check that case
> and processor cooling fans are operating. They have been known to
> fail!

I'd just like to pitch in again...

Overclocking your bus from 100 to 103 is not likely to exhibit problems.
Any problems you have were probably there at 100MHz also.

I'm talking about the folks who overclock their 66MHz (bus speed) Celeron's
to 100MHz.  Sure, there are Celerons that can make it and you won't see any
problems, but if that game you're playing exhibits a small calc error once
per hour, who'll notice?  You WILL notice it in Prime95 though.

Following along on what Brian Beesley is saying, the problem is definitely
heat.  The FPU in CPU's takes up a good 1/4 to 1/3 of the total die package,
but it's not normally used that much (although MMX does share the same
registers as the FPU).

When you play your games, run Word/Excel, you're just simply not generating
as much heat.  When you run an FPU intensive program, your CPU will really
heat up much more, and heat is the killer...raises resistance, causes
"misfires" with capacitive loads (e.g. memory, be it cache or registers),
timing problems, etc.

So, I urge anyone out there to not overclock their prime finding machines,
or if you're tempted, do so only moderately (100-103 bus speed for instance,
or possibly 66-75 is not extraordinary).

Those Abit motherboards with their "turbo" setting provides a nice, moderate
3% boost in clock speed and *I* would consider that safe anyway.

Aaron

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 08 Jun 1999 17:10:50 +0200
From: "Henk Stokhorst." <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: [First 10 million digit exponent]

L.S.,

> And below M36000000, there are 159,975 exponents (again repeating Brian)
> with at least 10 million digits.

Fine, but are the efforts being made in that region centrally registered?

YotN,

Henk Stokhorst.


________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 08 Jun 1999 10:55:06 -0400
From: George Woltman <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Mersenne Digest V1 #569

At 09:10 AM 6/8/99 +0100, you wrote:
>On Mon, Jun 07, 1999 at 04:37:32PM -0500, JON STRAYER wrote:
>> > You need to test exponent (10000000 / log 2) = 33219281 or above.
>
>Time for an n bit multiply using DWT is O(n log n) plus some fiddly
>bits to do with decreasing precision available when using longer
>convolutions - I'd probably estimate this as O(log log n).  As it
>takes n iterations for each test then to test an n digit number using
>the Lucas Lehmer test with DWT takes O(n^2 . log n . log log n).
>
>If we say that the current exponent is about 6.5 million then we need
>to test one which is approx 5 times bigger.  Using the above formula I
>make this 29.97 times longer for a 5 times increase in exponent size
>starting at 6.5 million.

My PII-400 is showing 1.121 and 1.236 seconds per iteration for FFT sizes
of 1.75M and 2M.  The crossover point will be somewhere in the vicinity
of 35,000,000.  Thus, the smallest 10 million digit Mersenne number
will take 33219281 * 1.121 seconds or 431 days.  In a few years when 2 GHz
chips are readily available this will be a more reasonable 80 days.



________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 08 Jun 1999 11:20:43 -0400
From: George Woltman <[EMAIL PROTECTED]>
Subject: Mersenne: Re: Self-test (was: Prime 95 Error Messages/ Misc)

Hi,

        An aside:  Please try to rename the subject like I did
above when the thread has strayed way off the original topic).  Thanks.

At 06:34 AM 6/8/99 -0500, Mikus Grinbergs wrote:
>I have seen opinions on this mailing list that the
>existing 16-hour self-test might not be enough to bring hardware
>deficiencies to light.  Someone suggested running the torture
>test - but is that more "rigorous" than the self-test?

It is more rigorous in that it can run for as long as you like,
whereas the full self-test takes "only" 16 hours if I recall correctly.

>And why
>should running the torture test for a week be any more definitive
>than already running Lucas-Lehmer on an assigned exponent?

The torture test is identical to a Lucas-Lehmer test except that
the end result is known and can be tested for correctness.
When you are done with a Lucas-Lehmer test your result will not be
proven right or wrong until a double-check or triple-check is done.

>My suggestion is that the QA people should "update" the built-in
>self-test to make it more likely to provoke hardware deficiencies.
>(Also, now that we are in the realm of "large" FFTs, is it still
>useful for the self-test to spend so much time on "short" FFTs ?)

I don't know how to make the self-test any more strenuous.  I suspect
short FFTs are just as hard on the CPU and cache as the large FFTs.
The large FFTs will access a little more memory.

Best regards,
George

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 08 Jun 1999 11:14:35 -0400
From: George Woltman <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Prime 95 Error Messages/ Misc

Hi,

At 04:39 PM 6/7/99 -0700, J. Williams wrote:
>Error: Illegal Sumout

This error can be a hardware problem, but is very often caused by a
faulty driver or program (usually related to the audio card).

>Is this anything I need to be concerned about and is there a listing of
>such messages?

This error and the others are briefly discussed in the readme.txt file.
Prime95 recovers from ILLEGAL SUMOUT ERRORs quite well, so there is
little to be concerned about.  If you get either of the other two errors,
then almost certainly you have a hardware problem, and a cause for concern.

>Also, a while ago the program reserved about 20 numbers for testing (all
>the way into the year 2001) then it and released them a few minutes later.
>Is this normal operation?

No.  It is possible the program misguessed your CPU speed (say it
thought you had a 2000 MHz Pentium), reserved a lot of exponents to
keep you busy, and then when your actual progress was less than
expected it returned the excess exponents.

>Any tips for optimization and usage?

No, there really is little you can do other than remove screen savers
and power save features that slow the CPU speed down.

>Finally, when and why does it communicate with the server (besides getting
>new numbers to test)?

Every 28 days it checks in to let the server know you are still 
working on the number.  It may also call in to let the server know
of any changes in the expected completion date.

Best regards,
George

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 8 Jun 1999 14:47:51 -0400
From: "Ernst W. Mayer" <[EMAIL PROTECTED]>
Subject: Mersenne: These go to 11 (WAS: blahblah...)

Paul Leyland <[EMAIL PROTECTED]> wrote:

>The radix is always 10.
{snip}
>or, more concisely, (1+1+1)^(1+1) + 1.
>
>Can anyone represent that number in fewer than (1+1+1)! ones?

How about

  1 << 1,

where the shift is, of course, decimal.

Your shifty friend,
- -Ernst
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 8 Jun 1999 19:49:03 +0100
From: [EMAIL PROTECTED]
Subject: Re: Mersenne: Re: [First 10 million digit exponent]

> > And below M36000000, there are 159,975 exponents (again repeating Brian)
> > with at least 10 million digits.
> 
> Fine, but are the efforts being made in that region centrally registered?

Not to my knowledge - seems pointless since few programs currently 
available can cope with exponents in the 30 millions. Prime95 v19 is 
still some way off, the early pre-pre-pre-release I have (as part of 
the QA team) won't accept exponents that big for factoring 
assignments, though it can attempt LL tests!

Of course it's very quick to generate lists of prime numbers in the 
30 millions by sieving. I stopped at 36 million because I thought 
that would yield more than enough work to start with!

Out of interest, I also wrote a quick-and-dirty program to check for 
small (< 2^32) factors of Mersenne numbers, ran that against my list 
of primes & eliminated 28,468 of the 159,975 candidates. (A 15-minute 
run on a PII-350). I need to check these & will post the files on my 
ftp server when I've done so.

Regards
Brian Beesley
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 8 Jun 1999 17:53:49 -0500 
From: "Griffith, Shaun" <[EMAIL PROTECTED]>
Subject: Mersenne: re: Island Theory...Settled?

Luke Welsh writes:

   TTBOMK, the theory was never formalized.  Regardless, Peter Lawrence
   Montgomery settled the issue:
       http://www2.netdoor.com/~acurry/mersenne/archive2/0032.html

I assume that the island theory was settled in the negative? I have read the
post, and others. Instead of settling the island theory, it seems that PLM
has supplied an alternative that fits and is well defined.

In http://www.silverlink.net/poke/archive4/0278.html
<http://www.silverlink.net/poke/archive4/0278.html>  I summarized the poor
man's predictor - MS Excel trendline of ln(Exponents of Mersenne Primes).

At the time, there were only 36 known Mersenne Primes. The trendline
predicted 

        p = 1.75(e^(0.4*Ordinal)),

so that trend(M38): p = 6987377.


Updating my spreadsheet to include the actual M36 & M37 gives 

        p = 1.77(e^(0.4*Ordinal)),

which predicts trend(M38): p = 7067233.

Note, however, that roundoff in the "0.4" multiplier causes a huge change in
the outcome. 
When I use the more accurate "0.391", the prediction is p = 5020196.

To me, this trendline seems about as good a fit as PLM's. The range of the
errors is about the same. The range of the squared errors is worse for PLM's
exponential distribution. The error is more centered using the log-linear
trendline.

I don't know enough about the rest, anyone have any comments here?

- -Shaun
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 8 Jun 1999 11:07:05 -0700
From: "Ethan Hansen" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Prime 95 Error Messages/ Misc

Aaron,

   You are correct in identifying overheating as the main problem.  Most
major manufactures give a maximum air temperature of 35-40C inside the PC
case.  If the air temperature is lower, the processor will run somewhat
faster.  The speedup is on the order of 1-3MHz per degree C, depending on
processor type.  None of this is to say that overclocking will not fail for
some other reason, but the 3% increase is not likely to cause problems
except in a very hot environment.

Regards,

Ethan

> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Aaron
> Blosser
> Sent: Tuesday, June 08, 1999 7:28 AM
> To: Mersenne@Base. Com
> Subject: RE: Mersenne: Prime 95 Error Messages/ Misc
>
>
> >
> > Note that simple overheating can also cause problems, even if your
> > system _isn't_ overclocked. Might be an idea to check that case
> > and processor cooling fans are operating. They have been known to
> > fail!
>
> I'd just like to pitch in again...
>
> Overclocking your bus from 100 to 103 is not likely to exhibit problems.
> Any problems you have were probably there at 100MHz also.
>
> I'm talking about the folks who overclock their 66MHz (bus speed)
> Celeron's
[SNIP]
>
> When you play your games, run Word/Excel, you're just simply not
> generating
> as much heat.  When you run an FPU intensive program, your CPU will really
> heat up much more, and heat is the killer...raises resistance, causes
> "misfires" with capacitive loads (e.g. memory, be it cache or registers),
> timing problems, etc.
>
> So, I urge anyone out there to not overclock their prime finding machines,
> or if you're tempted, do so only moderately (100-103 bus speed
> for instance,
> or possibly 66-75 is not extraordinary).
>
> Those Abit motherboards with their "turbo" setting provides a
> nice, moderate
> 3% boost in clock speed and *I* would consider that safe anyway.
>
> Aaron
>
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
>

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Tue, 08 Jun 1999 22:25:57 -0500
From: Gary Diehl <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Prime 95 Error Messages/ Misc

The system I built/overclocked is a Pentium-II 266 with a Shuttle
Computer International HOT-641 motherboard and 256 megs of 10ns RAM. 
The original bus clock speed was 66 mhz which I upped to 83 mhz in the
BIOS Configuration Utility.  I had to remove the case and train a 3"
desk fan on the CPU to keep it cool enough to run at 333 mhz but the 20%
increase in speed was worth it.  I ran a burn-in test on it for 2 solid
days when I first overclocked it to make sure I didn't have a weak
processor, and when I started Prime95 I tortured it for about a half of
a day with no errors so I figured it was good to go.  I think the
problem is heat-related, due to the onset of hotter months in this
climate and the poor ability of our air conditioner to handle the heat.

I have since started looking at P2-450mhz processors because adding an
extra 20% to the Prime # "expected completion date" test time after
having a shorter completion time is rather irksome.

Thanks to all the list members who offered different and vaild
viewpoints as to the nature of the problem.  I have since configured my
reported CPU speed in Prime95 back to 266 (which I forgot to do), and
the reported time between iterations increased to a more accurate
figure.

Gary Diehl



Brian J Beesley wrote:
> 

> 
> > It is also possible to have an overclocked CPU pass the full self test
> > suite, but later exhibit problems.  The likely culprit is simple wearout 
- --
> > the CPU initially was barely functional at the overclocked speed, but slowed
> > enough that it no longer runs.  If this happens, you usually can still run
> > the CPU at the rated speed.
> 
> Or, the errors are there all the time but at a low rate e.g. on
> average 1 every day. This is very likely to pass the 1 hour "self
> test" but will show up in actual use, or on the continuous "torture
> test" (see the "Options" menu in Prime95).
> 
> If you've overclocked your system at all (_not_ reccomended, but I
> know it can be successful in some cases) then I suggest you let
> the torture test run for a couple of days before committing yourself
> to doing "real" work. It can happen that an overclocked system
> appears to run fine for "office" applications but causes problems
> with Prime95 because very few other applications use the floating-
> point unit even half as intensively as Prime95 does.
> 
> Note that simple overheating can also cause problems, even if your
> system _isn't_ overclocked. Might be an idea to check that case
> and processor cooling fans are operating. They have been known to
> fail!
> 
> Finally (I think this is right - I'm sure George will chip in if not) there
> is a small but finite chance that you could get a very occasional
> "sum out error" even if your system is 100% perfect. This is due to
> abnormal combinations of data in the FFT triggering the "sanity
> check" in the code; the result may well be OK. The program should
> check that the "error" is "deterministic" (repeatable) rather than
> random and continue automatically if it is - though there will be an
> error log entry - PrimeNet uses this information to flag the result as
> "suspect", the exponent should then be re-assigned for an early
> double-check instead of waiting its turn as usual.
> 
> Regards
> Brian Beesley
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 09:48:34 GMT
From: "Brian J Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Self-test (was: Prime 95 Error Messages/ Misc)

George Woltman writes, in reply to Mikus Grinbergs:
 
> I don't know how to make the self-test any more strenuous.  I suspect
> short FFTs are just as hard on the CPU and cache as the large FFTs.
> The large FFTs will access a little more memory.

Most modern motherboards contain case and/or CPU temperature 
sensors which can be read by software. I believe the hardware 
interface is a de-facto standard, so portability shouldn't be a major 
concern. I suppose it would be possible to log the temperatures at 
which the tests were run & re-test if the temperature during a "live" 
run was found to be higher. But whether it's worth the effort is 
disputable - the same boards also have a high temperature alarm 
which will trigger if the temperature exceeds a limit (set in the 
BIOS) - my (Supermicro) board also comes with a Win 9x/NT utility 
(runs in the background, using about 0.5% CPU) which will sound 
"warnings" for high temperature, low fan speed, etc, etc. It has its 
own setup file, therefore does not require a reboot via BIOS Setup 
to change the settings.

When I built my dual PII system I did find I had a potential 
overheating problem, on Prime95 self-test one of the processors 
was triggerring the high temperature warning, nevertheless the 
reliability was OK. I fitted a case fan and the processor 
temperature dropped by about 10C, well out of the danger zone.
I get the impression that plastic cases with "tinfoil" screening need 
more assistance with ventilation than cases made of painted metal.

Incidentally the reason one processor runs hotter than the other is 
that the airflow in its vicinity is compromised by a bunch of 
HDD/FDD cables, so that processor fan tends to drag in "dirty" air.

If anyone has any ideas as to how the self-test could be made 
more rigorous, without taking forever, we'd sure like to hear them.


Regards
Brian Beesley
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 02:31:18 -0700
From: "Scott Kurowski" <[EMAIL PROTECTED]>
Subject: Mersenne: RE: Mersenne Digest V1 #570

Hi all,

(replying to the digest)

[Steinar:]
> >However we can't deny having seen cash prizes, in addition to
> >this satisfaction, make an observable difference in accelerating
> >Mersenne number research results.
>
> Can we? I'm sure you (Scott) is the one with most data on this...

I have one solid data point and a few weak ones.  The GIMPS newsletter
registration web form data has about 1,000 entries.  GIMPS participation is
apparently self-reported with prizes as a motivating factor for at least 16% of
registrants, though prizes alone did come in last:

26% good use of spare CPU time
18% just for fun or cool hobby
15% research participation
14% several reasons, in part for the prize
14% (no response entered)
06% help distributed computing
05% possible famous discovery
02% other (added to comments)
02% $50k prize money

The weaker evidence is from PrimeNet's account level responses.  First, the
April 1999 GIMPS Newsletter detailed both the v17 bug and the EFF prize - and
resulted in about 5% of recipients rejoining GIMPS and adding over 1,000 new
accounts the following few days.  The second account level response was EFF's
prize announcment finally gaining press (including CNN and Science), where
another 1,000 accounts joined in the span of a week.

In conclusion, I'm guessing people on the fence may be tipped by the extra
reason to play, but in retrospect my impression of prize effectiveness was more
subjective than I thought.

So while I'm on the subject, the list response regarding sponsoring a prize pool
amounted to 2 people offering $225 between them - I have my answer.


> - --Luke, who has such great vision that he lost a bet and
>         owes George a dinner (prime rib, of course)

That reminds me, during a lunch last summer it was Luke who made sure the prize
pool was all 1 digits to entertain all of you...! (I was too lazy to change all
the zeros on the web site.)  Luke is also an advisor to George and I, and helped
shape the PrimeNet status report content & layouts.


[John Williams:]
> Finally, when and why does it communicate with the server (besides getting
> new numbers to test)?

It will update the server at least every 28 days with when the exponents are
expected to finish.  This keeps your assignments 'alive' so PrimeNet will not
give them away after several months.  When there's a factoring or primality test
result, or you change your account settings, those are updated immediately.

Overall, the transactions are very fast and only a few hundred bytes each,
perhaps once a week on average.  The majority of the server's transactions are
assignment progress updates.


[Aaron Blosser:]
> I did notice in the prime.log that when errors occur, that info is sent to
> the Primenet server.
>
> On that note, Scott...what becomes of info sent indicating errors in the
> calculations?  Are those exponents flagged in someway, indicating that they
> are "suspect"?

PrimeNet completely ignores error messages and stuffs them into the results log
for George to analyze.  George requeues exponents with bad results through the
database synchronization process and his use of a few remote scheduling
commands.  Things seem to be working smoothly enough, but there's always room
for improvement.

Best regards,
scott


________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 08:17:54 -0400
From: "Pierre Abbat" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Self-test (was: Prime 95 Error Messages/ Misc)

> Most modern motherboards contain case and/or CPU temperature 
> sensors which can be read by software.

Is there a file in /proc that will tell me this?

phma
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 06:22:15 -0700
From: "Joth Tupper" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: These go to 11 (WAS: blahblah...)

Ground rules are critical, but how about

/.1

where "/" represents the APL-style monadic divide or multiplicative inverse.

1/.1

takes two.


- ----- Original Message -----
From: Ernst W. Mayer <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, June 08, 1999 11:47 AM
Subject: Mersenne: These go to 11 (WAS: blahblah...)


>
> Paul Leyland <[EMAIL PROTECTED]> wrote:
>
> >The radix is always 10.
> {snip}
> >or, more concisely, (1+1+1)^(1+1) + 1.
> >
> >Can anyone represent that number in fewer than (1+1+1)! ones?
>
> How about
>
>   1 << 1,
>
> where the shift is, of course, decimal.
>
> Your shifty friend,
> -Ernst
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
>

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 06:48:24 -0700 
From: Paul Leyland <[EMAIL PROTECTED]>
Subject: RE: Mersenne: These go to 11 (WAS: blahblah...)

> Ground rules are critical, but how about
> 
> /.1
> 
> where "/" represents the APL-style monadic divide or 
> multiplicative inverse.
> 
> 1/.1
> 
> takes two.

Indeed, but ".1" represents 1 / radix, so 1/.1 is just radix.

The whole point of my tongue-in-cheek posting was to indicate how an
implicit assumption of decimal notation had crept in, despite the clear
intention of the original to make the representation explicit.

So: you *cannot* assume any particular radix; your solution must work for
all integer radixes.


Paul
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 07:34:50 -0700
From: "John R Pierce" <[EMAIL PROTECTED]>
Subject: Mersenne: another CPU speed benchmark...

I just built up a Pentium-III 450 (128MB PC100 memory running at CAS/3), and
I'm testing it at 528MHz with a 115MHz system clock (rock stable so far).
Prime95 is benching M=7,636,483 [63 bits] at 0.190 seconds/iteration.  My
fastest box yet!  Is anyone still collecting CPU vs. prime95 speed
statistics?

- -jrp


________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 08:04:56 -0700 
From: Paul Leyland <[EMAIL PROTECTED]>
Subject: Mersenne: ECM factoring and P773

A small but valuable group of members of the Mersenne list have been
running the elliptic curve method to factor small Mersenne numbers,
with some impressive results so far.  The outstanding example is
Conrad Curry's discovery of the record-holding 53 digit factor of
M677, and still the only factor with more than fifty digits to be
found by ECM

I would like to request that one particular number be concentrated on.
That number is P773, also known as 2,773+ by the Cunningham project
and as 2^773+1 by everyone. Two small factors are known, 3 and 533371.
The composite cofactor has 227 digits.

The reason for this request is that it is a candidate for a SNFS
factorization (as are several others) and before embarking on a
large-scale computation, we want to be sure that it doesn't have a
small factor --- under fifty digits or so.

George's ECM page (http://www.mersenne.org/ecm.htm) records only the
work done by his prime95 program and reported to him.  Other work
in the past (I've done well over a thousand curves at B1=1M, for
instance) suggest that there are very probably no factors under 40
digits and that you should start with the B1 value set to 3 million. 
If you join in, please report progress back to George so that we can
move on to higher values of B1 when appropriate.

Prime95 is so efficient at factoring numbers of the form 2^n +/- 1
that non-Intel machines are at a significant disadvantage. 
Nonetheless, they are a valuable resource and owners of such might
find the ECMNET page
http://www.loria.fr/~zimmerma/records/ecmnet.html
useful.  I will be adding P773 to the ECMNET master server shortly.


Paul
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 09 Jun 1999 16:28:46 +0000
From: "David L. Nicol" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: EFF and 10,000,000 digits

Aaron Blosser wrote:
> 
> > I have heard some insider news that Intel *could* hit the 1 GigaHertz mark
> > by years end if they had a reason to

Did DEC not demonstrate a gigahertz Alpha chip shortly before Compaq 
purchased them?

________________________________________________________________________
  David Nicol 816.235.1187 UMKC Network Operations [EMAIL PROTECTED]
                          "unpersuasive and dubious"
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 09 Jun 1999 16:51:17 +0000
From: "David L. Nicol" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Self-test (was: Prime 95 Error Messages/ Misc)

Pierre Abbat wrote:
> 
> > Most modern motherboards contain case and/or CPU temperature
> > sensors which can be read by software.
> 
> Is there a file in /proc that will tell me this?
> 
> phma

This is the first I've heard of such sensors being a standard
item.  How long have they been a standard item?


________________________________________________________________________
  David Nicol 816.235.1187 UMKC Network Operations [EMAIL PROTECTED]
                "The radix is always 10." -- Paul Leyland
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 09 Jun 1999 10:42:20 -0700
From: Michael Gebis <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Self-test (was: Prime 95 Error Messages/ Misc) 

>>>>> "David" == David L Nicol <[EMAIL PROTECTED]>
>>>>> wrote the following on Wed, 09 Jun 1999 16:51:17 +0000

  David> Pierre Abbat wrote:
  >>  > Most modern motherboards contain case and/or CPU temperature >
  >> sensors which can be read by software.
  >> 
  >> Is there a file in /proc that will tell me this?
  >> 
  >> phma

  David> This is the first I've heard of such sensors being a standard
  David> item.  How long have they been a standard item?

Linux users will probably be interested in the lm_sensors package, which
is a kernel module that does what Pierre suggests: makes your system's
hardware health information available via the /proc filesystem.

The homepage is:
http://www.netroedge.com/~lm78/

Even non-linux users might be interested in these pages, as they contain
information about how "hardware health monitoring" works, a useful FAQ,
and plenty of useful related links.  (This is why I'm posting this in
response to David's question: the FAQ addresses his question.)

- --Mike

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 11:13:19 -0700
From: "Joth Tupper" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: EFF and 10,000,000 digits

Two things:

1) I do seem to recall a 1GHz Alpha announcement.

2) Was it Intel that bought the Alpha rights?  It might have been IBM but
was NOT Compac.

Joth

- ----- Original Message -----
From: David L. Nicol <[EMAIL PROTECTED]>
To: Aaron Blosser <[EMAIL PROTECTED]>
Cc: Mersenne@Base. Com <[EMAIL PROTECTED]>
Sent: Wednesday, June 09, 1999 9:28 AM
Subject: Re: Mersenne: EFF and 10,000,000 digits


> Aaron Blosser wrote:
> >
> > > I have heard some insider news that Intel *could* hit the 1 GigaHertz
mark
> > > by years end if they had a reason to
>
> Did DEC not demonstrate a gigahertz Alpha chip shortly before Compaq
> purchased them?
>
> ________________________________________________________________________
>   David Nicol 816.235.1187 UMKC Network Operations [EMAIL PROTECTED]
>                           "unpersuasive and dubious"
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
>

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 14:03:03 -0500 
From: "Willmore, David" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: EFF and 10,000,000 digits

Yes, you can find the presentation under their website at:
http://www.dec.com/alphaoem/microprocessorforum.htm

Cheers,
David

> ----------
> From:         David L. Nicol[SMTP:[EMAIL PROTECTED]]
> Sent:         Wednesday, June 09, 1999 11:28 AM
> To:   Aaron Blosser
> Cc:   Mersenne@Base. Com
> Subject:      Re: Mersenne: EFF and 10,000,000 digits
> 
> Aaron Blosser wrote:
> > 
> > > I have heard some insider news that Intel *could* hit the 1 GigaHertz
> mark
> > > by years end if they had a reason to
> 
> Did DEC not demonstrate a gigahertz Alpha chip shortly before Compaq 
> purchased them?
> 
> ________________________________________________________________________
>   David Nicol 816.235.1187 UMKC Network Operations [EMAIL PROTECTED]
>                           "unpersuasive and dubious"
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> 
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 14:05:42 -0500 
From: "Willmore, David" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Re: Self-test (was: Prime 95 Error Messages/ Misc)

For a few years.  Most PII class MBs have them and some late Socket7 ones
do, too.  They're not standard as far as how they interface both at the
hardware and software levels.  You need MB specific drivers.  Linux has a
project called lm_sensors (named after one of the early chips used to
perform this function, the National LM78).

Cheers,
David

> ----------
> From:         David L. Nicol[SMTP:[EMAIL PROTECTED]]
> Sent:         Wednesday, June 09, 1999 11:51 AM
> To:   Pierre Abbat; Yngvwe Mersenne
> Subject:      Re: Mersenne: Re: Self-test (was: Prime 95 Error Messages/
> Misc)
> 
> Pierre Abbat wrote:
> > 
> > > Most modern motherboards contain case and/or CPU temperature
> > > sensors which can be read by software.
> > 
> > Is there a file in /proc that will tell me this?
> > 
> > phma
> 
> This is the first I've heard of such sensors being a standard
> item.  How long have they been a standard item?
> 
> 
> ________________________________________________________________________
>   David Nicol 816.235.1187 UMKC Network Operations [EMAIL PROTECTED]
>                 "The radix is always 10." -- Paul Leyland
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> 
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 14:21:19 -0500 
From: "Willmore, David" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: EFF and 10,000,000 digits

1) see my other email.  Yes, they did.

2) Yes, it was Compaq.  Intel bought foundry technology that is used on the
StrongARM as well as the StrongARM archetecture itself.

Cheers,
David

> ----------
> From:         Joth Tupper[SMTP:[EMAIL PROTECTED]]
> Sent:         Wednesday, June 09, 1999 1:13 PM
> To:   GIMPS
> Subject:      Re: Mersenne: EFF and 10,000,000 digits
> 
> Two things:
> 
> 1) I do seem to recall a 1GHz Alpha announcement.
> 
> 2) Was it Intel that bought the Alpha rights?  It might have been IBM but
> was NOT Compac.
> 
> Joth
> 
> ----- Original Message -----
> From: David L. Nicol <[EMAIL PROTECTED]>
> To: Aaron Blosser <[EMAIL PROTECTED]>
> Cc: Mersenne@Base. Com <[EMAIL PROTECTED]>
> Sent: Wednesday, June 09, 1999 9:28 AM
> Subject: Re: Mersenne: EFF and 10,000,000 digits
> 
> 
> > Aaron Blosser wrote:
> > >
> > > > I have heard some insider news that Intel *could* hit the 1
> GigaHertz
> mark
> > > > by years end if they had a reason to
> >
> > Did DEC not demonstrate a gigahertz Alpha chip shortly before Compaq
> > purchased them?
> >
> > ________________________________________________________________________
> >   David Nicol 816.235.1187 UMKC Network Operations [EMAIL PROTECTED]
> >                           "unpersuasive and dubious"
> > ________________________________________________________________
> > Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> >
> 
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
> 
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

Date: Wed, 9 Jun 1999 15:33:46 -0400 (EDT)
From: Chip Lynch <[EMAIL PROTECTED]>
Subject: Re: Mersenne: EFF and 10,000,000 digits (fwd)

Intel worked out a licensing agreement with Digital just before Compaq
took over.  It gave them access to almost all of the technology of the
Alpha line, if I recall correctly.  I don't think anyone actually bought
the Alpha chips; the line stayed with Digital when it went to Compaq.  IBM
was not involved.

If I remember, Intel had better research into smaller data paths (so and
so many nanometers smaller than the Alpha line), but the Alphas had
superior clock rates... the combination of those two technologies (and a
million other miscellaneous marriages of technology) was suppossed to
produce killer chips.

Or something like that,
- ---Chip

       \\ ^ //
        (o o)
 ---oOO--(_)--OOo------------------------------------
| Chip Lynch            |   Computer Geek            |
| [EMAIL PROTECTED]       |                            | 
| (703) 465-4176   (w)  |   (202) 362-7978   (h)     |
 ----------------------------------------------------

- ---------- Forwarded message ----------
Date: Wed, 9 Jun 1999 11:13:19 -0700
From: Joth Tupper <[EMAIL PROTECTED]>
To: GIMPS <[EMAIL PROTECTED]>
Subject: Re: Mersenne: EFF and 10,000,000 digits

Two things:

1) I do seem to recall a 1GHz Alpha announcement.

2) Was it Intel that bought the Alpha rights?  It might have been IBM but
was NOT Compac.

Joth

- ----- Original Message -----
From: David L. Nicol <[EMAIL PROTECTED]>
To: Aaron Blosser <[EMAIL PROTECTED]>
Cc: Mersenne@Base. Com <[EMAIL PROTECTED]>
Sent: Wednesday, June 09, 1999 9:28 AM
Subject: Re: Mersenne: EFF and 10,000,000 digits


> Aaron Blosser wrote:
> >
> > > I have heard some insider news that Intel *could* hit the 1 GigaHertz
mark
> > > by years end if they had a reason to
>
> Did DEC not demonstrate a gigahertz Alpha chip shortly before Compaq
> purchased them?
>
> ________________________________________________________________________
>   David Nicol 816.235.1187 UMKC Network Operations [EMAIL PROTECTED]
>                           "unpersuasive and dubious"
> ________________________________________________________________
> Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
>

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm

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

End of Mersenne Digest V1 #571
******************************

Reply via email to