When will GIMPS's first prime...

[axn]

Originally Posted by axn

Well, we're at end of march, and it is at #4020. Most likely will fall off the list during next TdP (i.e. Feb 2019).

Nearing end of 2018, and it is at #4548. Still on track for Feb 2019 TdP race for it to finally drop off.

Originally Posted by CRGreathouse

Hmm. , and two to that power is about 22.5, so that's expected to occur when we're working with Mersenne exponents about 22.5 times larger. If we're around 30 million right now, that would be 675 million.

How do these numbers look, now that 4 have been found (assuming the latest find does pan out)? You can take the current wave front at 82m.

[CRGreathouse]

Originally Posted by CRGreathouse

Hmm. , and two to that power is about 22.5, so that's expected to occur when we're working with Mersenne exponents about 22.5 times larger. If we're around 30 million right now, that would be 675 million.

Assuming (for simplicity) that the work needed to test an exponent is proportional to the square of the exponent and that all prime exponents are tested, that would require about 10,000 times the total effort of GIMPS to date. If half of that effort happened in the past two years, and GIMPS' computing power doubles every two years (by some combination of Moore's law and recruitment), this would require 2 lg(ln(2) * 20,001) or 27.5 years. Assuming instead that it doubles every two years for a decade and then holds constant, it would take about 2 * 20000/2^5 + 10 = 1260 years. So the timeline depends strongly on the assumptions made.

Originally Posted by axn

How do these numbers look, now that 4 have been found (assuming the latest find does pan out)? You can take the current wave front at 82m.

We'd only need to go 4.743125... times further, to 389 million. Doubling every two years we'd get there mid-2023.

[axn]

Originally Posted by axn

Nearing end of 2018, and it is at #4548. Still on track for Feb 2019 TdP race for it to finally drop off.

Well, TdP came and went, I guess. Almost, but not quite -- now at 4955.

[axn]

Originally Posted by axn

Well, TdP came and went, I guess. Almost, but not quite -- now at 4955.

And it has reached #5000. One more and it is over.

[rudy235]

Originally Posted by axn

And it has reached #5000. One more and it is over.

Search Output Style:   HTML   Text   Text (printable)  
rank	prime	digits	who	when	comment
5001	21398269 - 1	420921	G1	Nov 1996	Mersenne 35 (**)
 	 	 	 	 	 
Used 0.1074 second(s) to find 1 primes matching the selection criteria: 
Comment includes 'mersenne 35'.

[mart_r]

22 years, 4 months and 3 days after the discovery.

Among the older predictions, sm88's from his last post seems to be closest to this day (Oct 30, 2018, which is off by 4 months and 17 days).
And remarkably, BQ's very first guess at the start of this thread comes just behind that (Aug 22, 2017, off by 1 year, 6 months and 25 days).

Edit: Generous interpretation also leaves a place for axn's prediction in post # 4, "5-8 years" after Dec 2010.

[CRGreathouse]

Originally Posted by axn

GPUs are the game changers. They will, in fact, keep pace with Moore's law, and bump the constants. I expect 2 decades (i.e before the end of 2030) for the necessary 8 primes to pop out.

I'd say I am very much on track to paying up on my $0.25 bet to axn. We've had 4 primes pop up in 9 years, with 11 years for the remaining 4. Not a guarantee by any means -- we've had a very good run this past decade -- but certainly my money wasn't as safe as I'd hoped.

For future reference: Mersenne 47 (M43112609) was the largest prime known at the time this thread was started; axn wins if Mersenne 55?? is discovered by Dec 31 2030.

[Bottom Quark]

Originally Posted by mart_r

22 years, 4 months and 3 days after the discovery.

Among the older predictions, sm88's from his last post seems to be closest to this day (Oct 30, 2018, which is off by 4 months and 17 days).
And remarkably, BQ's very first guess at the start of this thread comes just behind that (Aug 22, 2017, off by 1 year, 6 months and 25 days).

Haven't been on here in a while, but it was nice to see that my prediction was quite close.

I'll go ahead and make another one: there will be less than 55 known mersenne primes by the end of 2030, so M1398269 will still be on the top 20 list on 12/31/2030.