[QUOTE=delta_t;231329]Would it work if you set the client to the default assignment (GIMPS - what makes sense) and then use the Primenet site to set the appropriate core to the P-1 assignment?[/QUOTE]When it is changed in software, it sends that to the server and sets it. I checked that.

[QUOTE=Uncwilly;231327]<Lt. Scott>I cannot do it Capt'n.</scott>
[\QUOTE]

Bend the rules (my wife's favorite pass time):

Set RAM higher one evening; grab a bunch of P-1 assignments; set it lower again and hopefully it lets you keep them

Howcome it turns out that P-1 is worth doing before LL,
but only marginally so?

Is it simply down to choosing the bounds right?

David

[QUOTE=petrw1;231333]Bend the rules (my wife's favorite pass time):

Set RAM higher one evening; grab a bunch of P-1 assignments; set it lower again and hopefully it lets you keep them[/QUOTE]

Alternate: grab a bunch of P-1 assignments through the manual reservations page, add them to your worktodo.txt, and hopefully it lets you keep them. :smile:

[QUOTE=Mini-Geek;231347]Alternate: grab a bunch of P-1 assignments through the manual reservations page, add them to your worktodo.txt, and hopefully it lets you keep them. :smile:[/QUOTE]
I did that and it is.

[QUOTE=davieddy;231335]Howcome it turns out that P-1 is worth doing before LL,
but only marginally so?[/QUOTE]Not sure what you mean by "only marginally".

P-1 is always worth doing before L-L.

[quote]Is it simply down to choosing the bounds right?[/quote]Well, yes.

Very, very low bounds have such a low chance of finding a factor that they're not worth the overhead of starting. If one increases the bounds while looking at the corresponding chance of finding a factor plus the time needed to reach those bounds, at some point the percentage chance of finding a factor crosses the P-1's time-cost percentage of an LL run (or two), then drops back below it as bounds get higher.

If one graphs the factor-chance vs. time-cost curve, it always humps up to a peak, then declines. The optimum bounds to use are those which produce the peak value on that curve.

The bounds-choosing algorithm in the prime95 P-1 code chooses optimum bounds based on six (IIRC) different parameters. If "Available Memory" is low, it simply chooses to do only stage 1 of P-1, which uses about the same amount of memory as an L-L.

Paradox?
 

Hmm: food for thought.

This may or not be an oversimplification, but
let's agree that the probability of finding a factor increases with
the time spent on P-1.
But the rate of finding factors increases up to a certain time spent
per exponent then decreases again.
If our goal was simply to find as many factors as possible, we would
operate at the peak of the curve.
But for a given exponent, it is worth spending extra time as long as
the time<probability of factor*duration of 2 LLtests.

Of course it is more satisfactory to find a factor than merely prove a
number composite with 2 LLs, but for the purpose of finding primes, this
strategy would appear to be completely neutral.

Am I missing something?

Another consideration is that some computers are more
suited to factoring as opposed to LL than others.

David

[QUOTE=davieddy;231836]
But for a given exponent, it is worth spending extra time as long as
the time<probability of factor*duration of 2 LLtests.
[/QUOTE]

And that is exactly what Prime95 does.

[QUOTE=garo;231875]And that is exactly what Prime95 does.[/QUOTE]

Indeed. More info for how it figures cost vs benefit for TF and P-1 is on The Math page:
[url]http://www.mersenne.org/various/math.php[/url]

[QUOTE=garo;231875]And that is exactly what Prime95 does.[/QUOTE]
Or rather [I]would do[/I] if folk actually ran it:smile:

And more to the point, I raised the question
"Why the faffing about with doing P-1 before the last bit of TF?"
in the thread "TF: A job half done?"

David

This is probably a silly question, but does the "P-1/ECM stage 2 memory" refer to the allocation for each core or each processor? For example, I've heard that at least 256 MB of memory should be allowed for P-1 assignments. If I had a dual-core processor, should I enter 256 or 512 MB?