Mersenne Digest       Monday, February 19 2001       Volume 01 : Number 819




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

Date: Sat, 17 Feb 2001 20:06:44 -0500
From: Nathan Russell <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

On Sat, 17 Feb 2001 14:42:17 -0800, you wrote:

>This is mildly amusing.
>http://radified.com/Overclocking/oldest_overclocker.htm
>
>read the text below the picture, specifically, the 3rd paragraph :)
>
>-jrp

This reminded me of a discussion I had with a friend.  He mentioned he
used SETI:

"Ever looked into Prime95?"
"Yeah - I use that all the time."
"Why do you use SETI, then?  I've used SETI before, sure, but I really
prefer Prime95.  It's easier to find something with that, too."
"They aren't the same."
"Why not?"
"I thought Prime95 was just for overclocking and stuff."

And so it continued...

Frankly, it upsets me when I read a magazine article that mentions
Prime95 solely in the context of testing recently OCed machines.  

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

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

Date: Sun, 18 Feb 2001 07:02:09 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

On 17 Feb 2001, at 20:06, Nathan Russell wrote:

> This reminded me of a discussion I had with a friend.
> 
> "Ever looked into Prime95?"
[... snip ...]
> "I thought Prime95 was just for overclocking and stuff."
> 
> Frankly, it upsets me when I read a magazine article that mentions
> Prime95 solely in the context of testing recently OCed machines.  

I know what you mean ... but the fact is there's no such thing as 
_bad_ publicity.

However I do wonder how much of the recycling of expired exponents is 
due to people picking up Prime95 just to run a hardware test. The 
proportion of LL test assignments which complete seems to be dropping 
recently. I wonder if there is some problem where the program goes & 
picks up an assignment straight away on installation - would it be 
too much of a deterrent to normal users if it asked a straightforward 
question "Do you want to run a full self-test?" before anything else, 
& if so leave the setup of PrimeNet username etc. till later?

BTW I think it's a Darned Good Idea (tm) to run a full selftest on 
_any_ new system, and indeed periodically (say every year) on any 
system which is in continuing use, whether or not it's overclocked.


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 Feb 2001 02:24:36 -0600
From: Shane & Amy Sanford <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

I for one found my way to the Prime95 project through overclocking.  In the 
OCing community Prime95 has been considered the single best stress test 
tool available for some time now (especially when used in conjunction with 
a game that makes heavy use of the 3d side of the computer).  No doubt a 
high percent of these people get a exponent and then abandon it before 
completion but a few stick around and add significantly to the project.  A 
few minutes ago I calculated the teams I directly created which added up to 
2000+ cpu hours per day & would rank in the top 30 or so.  Not counting the 
3 or 4 OCing friends I know run Prime95 who add up to over 5000+ cpu hours 
per day.  Granted that is only a tiny tiny drop in the bucket but still my 
point is Prime95 has a huge presence in the OCing community and regardless 
of their motivation for running Prime95 their total combined contribution 
might come as a surprise.

BTW, Brian mentioned it seems the % of abandoned LL tests is on the rise as 
of late.  If this is actually the case does it possibly indicate that it's 
getting close to time raise the mhz threshold for LL vs. double check work 
assignment again?

Shane

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

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

Date: Sun, 18 Feb 2001 04:35:44 -0600
From: "Steve" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

>-----Original Message-----
>From: Brian J. Beesley
<snip>
>However I do wonder how much of the recycling of expired exponents is
>due to people picking up Prime95 just to run a hardware test. The
>proportion of LL test assignments which complete seems to be dropping
>recently. I wonder if there is some problem where the program goes &
>picks up an assignment straight away on installation - would it be
>too much of a deterrent to normal users if it asked a straightforward
>question "Do you want to run a full self-test?" before anything else,
>& if so leave the setup of PrimeNet username etc. till later?

I agree with Brian how that would help to cut down on abandoned assignments,
but as we know, some of those are not abandoned but just late reporting in.
Which brings up a situation I've just come across which I believe could be
easily fixed.

A couple of weeks ago one of our PCs (14 days before completing current job)
was assigned an exponent in the 10M range which obviously had been assigned
to someone else earlier. No problem, that happens all the time. A few days
later the original assignee finished the exponent and it dropped off our
account report. Again no problem, that happens sometimes too. Then a couple
of days ago our PC finished its exponent and reported it in. Since it hadn't
been 28 days since last checkin, it just went ahead and started on the
exponent which someone else had already finished. This PC dropped off our
account report altogether, reducing the number of "machines assigned to
primenet". Now it has happened before that a reassigned exponent was
finished by the original owner AFTER one of our machines started testing it
and I realize there's not much that can be done about that. However, how
much trouble would it be to have it check with the server before starting on
a new exponent to make sure it hadn't already been finished by someone else?
It had to communicate with the server anyway at that point to report the
completed assignment.

Yeah I know, no big deal, it will be recorded as a double-check when ours
finishes. (At least I think it will.) Just seems like it would have been
easy enough to prevent happening in the first place.

Steve Harris



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

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

Date: Sun, 18 Feb 2001 12:22:49 +0100
From: "Martijn Kruithof" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

Actually as long as you have not started you assignment and you do
intermediairy check ins of status, and the assignment is finished before you
started working on it, it will be dropped from your assignment list, it only
stays if you are already working on it.


Martijn
- ----- Original Message -----
From: "Steve" <[EMAIL PROTECTED]>
To: "Brian J. Beesley" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Sunday, February 18, 2001 11:35 AM
Subject: Re: Mersenne: hah


>
> >-----Original Message-----
> >From: Brian J. Beesley
> <snip>
> >However I do wonder how much of the recycling of expired exponents is
> >due to people picking up Prime95 just to run a hardware test. The
> >proportion of LL test assignments which complete seems to be dropping
> >recently. I wonder if there is some problem where the program goes &
> >picks up an assignment straight away on installation - would it be
> >too much of a deterrent to normal users if it asked a straightforward
> >question "Do you want to run a full self-test?" before anything else,
> >& if so leave the setup of PrimeNet username etc. till later?
>
> I agree with Brian how that would help to cut down on abandoned
assignments,
> but as we know, some of those are not abandoned but just late reporting
in.
> Which brings up a situation I've just come across which I believe could be
> easily fixed.
>
> A couple of weeks ago one of our PCs (14 days before completing current
job)
> was assigned an exponent in the 10M range which obviously had been
assigned
> to someone else earlier. No problem, that happens all the time. A few days
> later the original assignee finished the exponent and it dropped off our
> account report. Again no problem, that happens sometimes too. Then a
couple
> of days ago our PC finished its exponent and reported it in. Since it
hadn't
> been 28 days since last checkin, it just went ahead and started on the
> exponent which someone else had already finished. This PC dropped off our
> account report altogether, reducing the number of "machines assigned to
> primenet". Now it has happened before that a reassigned exponent was
> finished by the original owner AFTER one of our machines started testing
it
> and I realize there's not much that can be done about that. However, how
> much trouble would it be to have it check with the server before starting
on
> a new exponent to make sure it hadn't already been finished by someone
else?
> It had to communicate with the server anyway at that point to report the
> completed assignment.
>
> Yeah I know, no big deal, it will be recorded as a double-check when ours
> finishes. (At least I think it will.) Just seems like it would have been
> easy enough to prevent happening in the first place.
>
> Steve Harris
>
>
>
> _________________________________________________________________________
> 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 Feb 2001 08:36:32 -0600
From: "Steve" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

Depends on what you mean by "assignment list". As soon as the first person
finishes the exponent it is dropped from your account report list, whether
you have started on it or not. But it is not removed from your 'worktodo'
file. Reporting your completion of another exponent is NOT treated as
"checking in". Therefore it will (and has) started work on a previously
completed assignment. You are correct (I believe) that it will drop from the
"worktodo" list if you do check in before it starts, but if you only check
in every 28 days there's an even chance that it won't catch it in time. If
it checked in when it was about to start (that is, when it reported
finishing the previous one) then it would catch it every time.

Steve Harris

- -----Original Message-----
From: Martijn Kruithof <[EMAIL PROTECTED]>
To: Steve <[EMAIL PROTECTED]>
Cc: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Date: Sunday, February 18, 2001 5:24 AM
Subject: Re: Mersenne: hah


>Actually as long as you have not started you assignment and you do
>intermediairy check ins of status, and the assignment is finished before
you
>started working on it, it will be dropped from your assignment list, it
only
>stays if you are already working on it.
>
>
>Martijn
>----- Original Message -----
>From: "Steve" <[EMAIL PROTECTED]>
>To: "Brian J. Beesley" <[EMAIL PROTECTED]>
>Cc: <[EMAIL PROTECTED]>
>Sent: Sunday, February 18, 2001 11:35 AM
>Subject: Re: Mersenne: hah
>
>
>>
>> >-----Original Message-----
>> >From: Brian J. Beesley
>> <snip>
>> >However I do wonder how much of the recycling of expired exponents is
>> >due to people picking up Prime95 just to run a hardware test. The
>> >proportion of LL test assignments which complete seems to be dropping
>> >recently. I wonder if there is some problem where the program goes &
>> >picks up an assignment straight away on installation - would it be
>> >too much of a deterrent to normal users if it asked a straightforward
>> >question "Do you want to run a full self-test?" before anything else,
>> >& if so leave the setup of PrimeNet username etc. till later?
>>
>> I agree with Brian how that would help to cut down on abandoned
>assignments,
>> but as we know, some of those are not abandoned but just late reporting
>in.
>> Which brings up a situation I've just come across which I believe could
be
>> easily fixed.
>>
>> A couple of weeks ago one of our PCs (14 days before completing current
>job)
>> was assigned an exponent in the 10M range which obviously had been
>assigned
>> to someone else earlier. No problem, that happens all the time. A few
days
>> later the original assignee finished the exponent and it dropped off our
>> account report. Again no problem, that happens sometimes too. Then a
>couple
>> of days ago our PC finished its exponent and reported it in. Since it
>hadn't
>> been 28 days since last checkin, it just went ahead and started on the
>> exponent which someone else had already finished. This PC dropped off our
>> account report altogether, reducing the number of "machines assigned to
>> primenet". Now it has happened before that a reassigned exponent was
>> finished by the original owner AFTER one of our machines started testing
>it
>> and I realize there's not much that can be done about that. However, how
>> much trouble would it be to have it check with the server before starting
>on
>> a new exponent to make sure it hadn't already been finished by someone
>else?
>> It had to communicate with the server anyway at that point to report the
>> completed assignment.
>>
>> Yeah I know, no big deal, it will be recorded as a double-check when ours
>> finishes. (At least I think it will.) Just seems like it would have been
>> easy enough to prevent happening in the first place.
>>
>> Steve Harris
>>
>>
>>
>> _________________________________________________________________________
>> 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 Feb 2001 10:25:14 -0500
From: Nathan Russell <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

On Sun, 18 Feb 2001 07:02:09 -0000, Brian Beesley wrote:


>I know what you mean ... but the fact is there's no such thing as 
>_bad_ publicity.

True enough.  

>However I do wonder how much of the recycling of expired exponents is 
>due to people picking up Prime95 just to run a hardware test. The 
>proportion of LL test assignments which complete seems to be dropping 
>recently. I wonder if there is some problem where the program goes & 
>picks up an assignment straight away on installation - would it be 
>too much of a deterrent to normal users if it asked a straightforward 
>question "Do you want to run a full self-test?" before anything else, 
>& if so leave the setup of PrimeNet username etc. till later?

That might not be a bad idea - especially since IIRC the first thing
done after setup is to request exponents.  

>BTW I think it's a Darned Good Idea (tm) to run a full selftest on 
>_any_ new system, and indeed periodically (say every year) on any 
>system which is in continuing use, whether or not it's overclocked.

Agreed.  

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

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

Date: Sun, 18 Feb 2001 10:27:23 -0500
From: Nathan Russell <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

On Sun, 18 Feb 2001 02:24:36 -0600, you wrote:

>I for one found my way to the Prime95 project through overclocking.  In the 
>OCing community Prime95 has been considered the single best stress test 
>tool available for some time now (especially when used in conjunction with 
>a game that makes heavy use of the 3d side of the computer).  No doubt a 
>high percent of these people get a exponent and then abandon it before 
>completion but a few stick around and add significantly to the project.  A 
>few minutes ago I calculated the teams I directly created which added up to 
>2000+ cpu hours per day & would rank in the top 30 or so.  Not counting the 
>3 or 4 OCing friends I know run Prime95 who add up to over 5000+ cpu hours 
>per day.  Granted that is only a tiny tiny drop in the bucket but still my 
>point is Prime95 has a huge presence in the OCing community and regardless 
>of their motivation for running Prime95 their total combined contribution 
>might come as a surprise.

Agreed - and congrats on your contribution.  I wasn't trying to upset
overclockers in general.  I simply meant to say that, perhaps, the
program should go further out of its encourage people to properly quit
the project.  

>BTW, Brian mentioned it seems the % of abandoned LL tests is on the rise as 
>of late.  If this is actually the case does it possibly indicate that it's 
>getting close to time raise the mhz threshold for LL vs. double check work 
>assignment again?

George just updated it with the launch of v20, so I wouldn't say so.
Certainly, there's room for discussion.  

>Shane

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

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

Date: Sun, 18 Feb 2001 17:49:53 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

On 18 Feb 2001, at 2:24, Shane & Amy Sanford wrote:

> I for one found my way to the Prime95 project through overclocking.  

As I said, there's no such thing as _bad_ publicity! I actually came 
across Prime95 whilst looking for a benchmarking program to see if my 
266 MHz PII system was performing as it should be.

>   In the 
> OCing community Prime95 has been considered the single best stress test 
> tool available for some time now (especially when used in conjunction with 
> a game that makes heavy use of the 3d side of the computer).  No doubt a 
> high percent of these people get a exponent and then abandon it before 
> completion but a few stick around and add significantly to the project.  

Sure. I really don't have a problem with that. My point was simply 
that if the options were set up a bit differently, more of those who 
never had any intention of completing a test might not set up in a 
way which causes them to actually take an assignment - which 
effectively ties the assignment up for (by default) 88 days. This in 
itself is not a problem, but the extra transactions on the PrimeNet 
server caused by this "waste" might be, one day.

What really does hurt is people starting a test, getting a 
substantial way through & then abandoning.

Another suggestion I might make is that the first assignment given to 
every new user/system should be a doublecheck. (Unless they 
specifically override the assignment type in the Test menu). This 
would complete fast enough to encourage new users by getting their 
name onto the league table, & might encourage more of the "casual" 
users to stay with the project.

> If this is actually the case does it possibly indicate that it's 
> getting close to time raise the mhz threshold for LL vs. double check work 
> assignment again?

No doubt opinions will differ on this. If you bump the threshold then 
you're signalling to a percentage of contributers that their system 
is no longer "good enough" and you might lose their contribution 
altogether.

My feeling is that we should be clearing LL & DC assignments at about 
the same rate. If the threshold stays the same then DC assignments 
will start to lag behind as new systems tend to be fast, whilst 
older, slower systems are retired or upgraded. Increasing the 
threshold will switch some of the slower systems from LL assignments 
to DC assignments, thus restoring the balance.

Another way to look at the balance is to examine the PrimeNet status 
report at 0800 UTC. There tends to be a balance of LL assignments 
which were recycled at 0600 still to be reassigned, whereas all the 
recycled DC assignments have usually gone by then. The indication 
here is that DC work is popular - perhaps we shouldn't be too keen to 
up the threshold whilst this state of affairs persists.

Yet another way is to compare the time taken for a DC assignment on a 
"threshold" system to the time taken for a LL test on a "state of the 
art" system. At present a typical DC assignment on a 300 MHz PII 
takes about half as long to run as a typical LL assignment on a 1.2 
GHz Athlon. Personally I don't feel this is unreasonable.


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 Feb 2001 13:31:50 -0500
From: Jud McCranie <[EMAIL PROTECTED]>
Subject: Mersenne: [Mersenne] WorkToDo file

I've edited my WorkToDo file to put the tests in order so that the ones 
that take the least time will be done first.  Will my doing this mess 
anything up?

+--------------------------------------------------------+
|                  Jud McCranie                          |
|                                                        |
| 137*2^261147+1 is prime!  (78,616 digits, 5/2/00)      |
+--------------------------------------------------------+

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

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

Date: Sun, 18 Feb 2001 14:04:08 -0500
From: Nathan Russell <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

On Sun, 18 Feb 2001 17:49:53 -0000, Brian J. Beesley wrote:

>On 18 Feb 2001, at 2:24, Shane & Amy Sanford wrote:
>>   In the 
>> OCing community Prime95 has been considered the single best stress test 
>> tool available for some time now (especially when used in conjunction with 
>> a game that makes heavy use of the 3d side of the computer).  No doubt a 
>> high percent of these people get a exponent and then abandon it before 
>> completion but a few stick around and add significantly to the project.  
>
>Sure. I really don't have a problem with that. My point was simply 
>that if the options were set up a bit differently, more of those who 
>never had any intention of completing a test might not set up in a 
>way which causes them to actually take an assignment - which 
>effectively ties the assignment up for (by default) 88 days. This in 
>itself is not a problem, but the extra transactions on the PrimeNet 
>server caused by this "waste" might be, one day.

Indeed; this would also be more convenient for the overclockers, since
they could immediately start a self-test.  

>What really does hurt is people starting a test, getting a 
>substantial way through & then abandoning.

True.  One of the problems GIMPS faces, compared with other
distributed computing projects, is that our 'work units' take weeks or
months; if people joined, e.g., distributed.net for a one-day
self-test, they'd tie up a day's worth of work, and return most of it.

>Another suggestion I might make is that the first assignment given to 
>every new user/system should be a doublecheck. (Unless they 
>specifically override the assignment type in the Test menu). This 
>would complete fast enough to encourage new users by getting their 
>name onto the league table, & might encourage more of the "casual" 
>users to stay with the project.

On the other hand, some users might become upset at being assigned a
double-check, and not realize that they could change the setting in
question.

>My feeling is that we should be clearing LL & DC assignments at about 
>the same rate. If the threshold stays the same then DC assignments 
>will start to lag behind as new systems tend to be fast, whilst 
>older, slower systems are retired or upgraded. Increasing the 
>threshold will switch some of the slower systems from LL assignments 
>to DC assignments, thus restoring the balance.

Or course, there's no absolute rule that we must always maintain such
a balance - if LL assignments start building a backlog, people will
become upset at the length of time those assignments are taking, and
will then switch to DC.  

>Another way to look at the balance is to examine the PrimeNet status 
>report at 0800 UTC. There tends to be a balance of LL assignments 
>which were recycled at 0600 still to be reassigned, whereas all the 
>recycled DC assignments have usually gone by then. 

Of course, many people (particularly members of this list - myself
included) go out of their way to claim the smaller DC assignments when
they are reissued.  To be sure, I've done the same with LL before, but
frankly not as often.  

>Yet another way is to compare the time taken for a DC assignment on a 
>"threshold" system to the time taken for a LL test on a "state of the 
>art" system. At present a typical DC assignment on a 300 MHz PII 
>takes about half as long to run as a typical LL assignment on a 1.2 
>GHz Athlon. Personally I don't feel this is unreasonable.

Of course, the 1.2 Ghx Athlon is right at the top of the "state of the
art", while systems well slower than a PII are still usable.  

>Regards
>Brian Beesley

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

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

Date: Sun, 18 Feb 2001 14:20:29 -0500
From: Nathan Russell <[EMAIL PROTECTED]>
Subject: Re: Mersenne: [Mersenne] WorkToDo file

On Sun, 18 Feb 2001 13:31:50 -0500, Jud McCranie wrote:

>I've edited my WorkToDo file to put the tests in order so that the ones 
>that take the least time will be done first.  Will my doing this mess 
>anything up?

Not to my knowledge.  The one thing to be careful of, however, is that
if you have a very long test presently coming after the small ones,
and it is moved to the front (the opposite of what you're doing), it
may be large enough to push the others off the edge of the amount of
work you're set to queue

Assignments set to start beyond that time (with a grace period, the
length of which I can't recall) will be returned to the server.  

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

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

Date: Sun, 18 Feb 2001 22:10:33 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: hah

On 18 Feb 2001, at 14:04, Nathan Russell wrote:

> >Another suggestion I might make is that the first assignment given to 
> >every new user/system should be a doublecheck. [... snip ...]
> 
> On the other hand, some users might become upset at being assigned a
> double-check, and not realize that they could change the setting in
> question.

If we decide to change this, the fact should be well documented. We 
could argue that completing a DC assignment is a sort of "initiation 
test". Serious new users with fast systems will not be delayed long; 
in any case, they could drop the assignment easily enough, or pre-
select LL testing before the first connection to PrimeNet.
> 
> >My feeling is that we should be clearing LL & DC assignments at about 
> >the same rate. [... snip ...]
> 
> Or course, there's no absolute rule that we must always maintain such
> a balance - if LL assignments start building a backlog, people will
> become upset at the length of time those assignments are taking, and
> will then switch to DC.  

No, there's no _rule_, and I'm not implying that there should be. But 
I feel that keeping some sort of a balance is a reasonable objective.
> 
> >Another way to look at the balance is to examine the PrimeNet status 
> >report at 0800 UTC. There tends to be a balance of LL assignments 
> >which were recycled at 0600 still to be reassigned, whereas all the 
> >recycled DC assignments have usually gone by then. 
> 
> Of course, many people (particularly members of this list - myself
> included) go out of their way to claim the smaller DC assignments when
> they are reissued.  To be sure, I've done the same with LL before, but
> frankly not as often.  

I've three "slower" systems running LL tests (a dual PII-350 and a 
Celeron 366) and I try to find smaller LL assignments for those. I've 
recently retired a couple of P100 systems. Not that they're totally 
useless, just that I think I can find better things to do with them.
> 
> >Yet another way is to compare the time taken for a DC assignment on a 
> >"threshold" system to the time taken for a LL test on a "state of the 
> >art" system. At present a typical DC assignment on a 300 MHz PII 
> >takes about half as long to run as a typical LL assignment on a 1.2 
> >GHz Athlon. Personally I don't feel this is unreasonable.
> 
> Of course, the 1.2 Ghx Athlon is right at the top of the "state of the
> art", while systems well slower than a PII are still usable.  

Naturally I agree. (Though the 1.2 GHz Athlon undoubtedly will be 
toppled from its perch; I'd guess sooner rather than later!) What I 
meant by "threshold" system is one at the cut-off point between 
getting LL & DC assignments unless the user overrides the default 
assignment type. Which is, I believe, 300 MHz at the moment.

I feel that if the time taken to run a DC assignment on a "threshold" 
system (meaning as indicated) starts to approach the time to run a LL 
assignment on a SoA system, perhaps it's time to think about moving 
the threshold up.

BTW the three indicators of DC/LL balance I indicated do not 
neccessarily point in the same direction; indeed they may well be 
contradictory to some extent. I see this as an indication that the 
cut-off point is not too far wrong at the moment, though it would not 
upset me personally if it were to move up to 350 MHz or even 400 MHz.


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 Feb 2001 21:39:45 -0500
From: "Joshua Zelinsky" <[EMAIL PROTECTED]>
Subject: Mersenne: Accuracy of completion dates.

Suggestion: the next version of Prime95 should contact the server after 
every 10% of an LL test. Based on how long this took, the server would 
calculate the probable finishing time. This estimate would probably be more 
accurate than the rough estimate based on how many hours the computer is 
estimated to be on which ignores for what the computer is normally used. The 
only drawback I see is the strain on the server.

Regards,

Joshua Zelinsky

_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com

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

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

Date: Mon, 19 Feb 2001 19:31:56 +1300
From: "Halliday, Ian" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Accuracy of completion dates.

This is a cryptographically signed message in MIME format.

- --------------ms391FCE4B6556D2A10B367471
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Joshua Zelinsky wrote:
> 
> Suggestion: the next version of Prime95 should contact the server after
> every 10% of an LL test. Based on how long this took, the server would
> calculate the probable finishing time. 

I disagree. I have an old but still half serviceable machine which is
still doing double checks. It's a P100 which sits in a cupboard and I
only connect it to the net to return results and get new exponents. Its
CD drive doesn't work any longer, its speakers don't work any longer,
its mouse doesn't work any longer and I have to find a modem cable each
month to connect it. Win95 without a mouse isn't too much fun. This PC
won't be connecting to the net every 10% but it's still in enough of
working order to return about one double check each month or so. When it
dies or isn't welcome to do prime95 work in this style any more, it will
retire. I suspect that there are other people with old pcs in cupboards
doing nothing but prime95 work. I hope so - I'd hate to think I was the
only one.

Regards,

Ian
- --
Ian W Halliday, BA Hons, MIMIS, AAIBF Snr, ATMB, CL
+64 25 245 6089
http://baptism.co.nz
Excel in all we do
- --------------ms391FCE4B6556D2A10B367471
Content-Type: application/x-pkcs7-signature; name="smime.p7s"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="smime.p7s"
Content-Description: S/MIME Cryptographic Signature
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- --------------ms391FCE4B6556D2A10B367471--

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

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

Date: Mon, 19 Feb 2001 00:56:58 -0600
From: Ken Kriesel <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Processor short family histories

At 09:44 AM 2/13/2001 -0500, Jud McCranie wrote:
>At 12:54 AM 2/13/2001 -0600, Ken Kriesel wrote:
>
>>Intel offered the 286 with 6, 8, 10, and 12.5 Mhz on one data sheet.
>>AMD got to 16 on this one, but an early data sheet lists 4, 6, and 8
>>(and says reprinted by permission of Intel).  FPU was separate.
>>I don't recall a 286-20.
>
>Dell had one.  At the time I got my Dell 20 MHz 386 (fall 1987) they had a 
>20 MHz 286.
>
> > The 386 debuted at 12.5 and 16 Mhz.
>
>I thought it debuted at 16.  I never heard of a 12.5 MHz 386.

12.5 and 16 is what the preliminary data sheet said when Intel
first printed & distributed them in 1985.  I believe Zenith made
12Mhz machines.  The manufacturers who could get 16Mhz
chips used the higher speed, naturally.


Ken

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

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

Date: Mon, 19 Feb 2001 19:44:32 -0000
From: "Brian J. Beesley" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Accuracy of completion dates.

On 18 Feb 2001, at 21:39, Joshua Zelinsky wrote:

> Suggestion: the next version of Prime95 should contact the server after 
> every 10% of an LL test. Based on how long this took, the server would 
> calculate the probable finishing time. This estimate would probably be more 
> accurate than the rough estimate based on how many hours the computer is 
> estimated to be on which ignores for what the computer is normally used. The 
> only drawback I see is the strain on the server.

This is sort of feasible but I still prefer my earlier suggestion 
that new users should (by default) run a DC assignment first.


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 Feb 2001 23:02:46 -0500
From: Jeff Woods <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Accuracy of completion dates.

At 09:39 PM 2/18/01 -0500, you wrote:

>Suggestion: the next version of Prime95 should contact the server after 
>every 10% of an LL test. Based on how long this took, the server would 
>calculate the probable finishing time. This estimate would probably be 
>more accurate than the rough estimate based on how many hours the computer 
>is estimated to be on which ignores for what the computer is normally 
>used. The only drawback I see is the strain on the server.

It can already be CONFIGURED to do so, not on percentages, but on number of 
days between contacts.   I have mine set to report daily, so my individual 
report is always up to date.
_________________________________________________________________________
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 #819
******************************

Reply via email to