Factoring bit depth?

[davieddy]

Originally Posted by R.D. Silverman

Since when is an algorithm for finding Mersenne primes "part of the real
world".?? None of what we are doing is remotely related to modelling the
real world.

Models can be wrong. The one I presented for the expected time to
test a Mersenne candidate for primality is not. [I never actually wrote down
the objective function, but it follows easily from the definitions that I
gave] If you believe that it is, please tell us what is wrong.

Does your "model" resemble mine:
7% more exponents tested per year (20% more 'puting) giving
George a vacation every 6 years?

Or were you thinking about Marilyn Monroe?

David

Resident somethingorother

[davieddy]

Originally Posted by chalsall

There are persons who are known for thinking...

There are persons who are known for doing...

There are persons who are known for thinking and doing...

Thank your god(s) for all....

George Bernard Shaw:
"Those who can do. Those who can't teach."
Anon: "Those who can't teach, teach teachers"

Graffitti in the Gents at the "Kosher Horses":
Philosopher A: "To do is to be"
Philosopher B: "To be is to do"
Frank Sinatra: "Doo be doo be doo"

David

[science_man_88]

Originally Posted by davieddy

Anon: "Those who can't teach, teach teachers"

I was told I tried teaching the teachers in grade 8, but I was testing grade 9 students for their chemistry test on the periodic table.

[LaurV]

Originally Posted by davieddy

Frank Sinatra: "Doo be doo be doo"
David

Fred Flintstone: Ya be da be doo! (well, not exact, but more thematic)

[cmd]

Originally Posted by LaurV

Fred Flintstone: Ya be da be doo! (well, not exact, but more thematic)

0=((ob|adi),(ob|ada))



;)C0onway

[davieddy]

My instinct is that it comes down to "turnover rate".

Think Henry Ford and "Time and Motion".

71 bits takes such a short time that folk reserve
a large number at a time.
Some then apparently feel no obligation to actually do the needful
and simply sit on them.

I think my idea of 72 bits as "one assignment" makes sense
from any point of view.

David

[TheJudger]

Hi!

David: what do you think is a good turnover time for TF assignments?
No matter of the runtime of current assignments I try to receive work for one week at once.

Oliver

[davieddy]

Originally Posted by TheJudger

Hi!

David: what do you think is a good turnover time for TF assignments?
No matter of the runtime of current assignments I try to receive work for one week at once.

Oliver

I would have thought that (like many GIMPS things) it would not be
critical. Bob may have other ideas!

My point was that those of us who have thought about
it think 72 bits for the 53M range LL assignments should be comfortably
achievable, (and hopefully with quite a bit to spare).

From what I've seen so far, hardly any exponents are getting dished
out with more than 70 bits ATM.
No wonder 80% get expired/returned half-chewed.

The next Mersenne prime is expected to need a lot of searching
for, and hence some dedication on the part of those engaged in the search.
Work at the coalface (53M) is not for the faint-hearted!

David

PS Does anyone still think a 100M digit prime will be found before
2025?

PPS If I avoided answering your question directly, it was
probably because you are not exactly the "average"
GIMPS participant

[drh]

Originally Posted by davieddy

I would have thought that (like many GIMPS things) it would not be
critical. Bob may have other ideas!

My point was that those of us who have thought about
it think 72 bits for the 53M range LL assignments should be comfortably
achievable, (and hopefully with quite a bit to spare).

From what I've seen so far, hardly any exponents are getting dished
out with more than 70 bits ATM.
No wonder 80% get expired/returned half-chewed.

The next Mersenne prime is expected to need a lot of searching
for, and hence some dedication on the part of those engaged in the search.
Work at the coalface (53M) is not for the faint-hearted!

David

PS Does anyone still think a 100M digit prime will be found before
2025?

PPS If I avoided answering your question directly, it was
probably because you are not exactly the "average"
GIMPS participant

David/Oliver

I reserve similiar to Oliver, a few days, up to 2 weeks worth at a time, but as long as I'm not traveling or have Internet access, I upload my results every night. Maybe that's a little more than I need to, but they do get dished out immediately afterward ... just trying to keep things moving a little bit.

Doug

[chalsall]

Originally Posted by davieddy

PS Does anyone still think a 100M digit prime will be found before 2025?

Yes. I do.

Why? Because the work in quantum computing is making great progress. (Gotta love those q-bits!)

My prediction (out of my ass): by 2020.

[davieddy]

Originally Posted by chalsall

Yes. I do.

Why? Because the work in quantum computing is making great progress. (Gotta love those q-bits!)

My prediction (out of my ass): by 2020.

Like nuclear fusion?

Leave the art of anal articulation to me