Merry Christmas and a prime! (M50 related)

[ATH]

Unfortunately 77232917 = 5 (mod 8) so there will be no primitive trinomials:

https://maths-people.anu.edu.au/~brent/trinom.html

https://maths-people.anu.edu.au/~brent/trinom-old.html

https://arxiv.org/abs/1005.1967

[heliosh]

Oooh :)

How are the chances?

[CRGreathouse]

Originally Posted by heliosh

Attachment 17467
How are the chances?

2^77232917-1 has 23249425 decimal digits. The expected number of instances of ten consecutive identical decimal digits is about (23249425-9)/10^9 = 0.023249416. The probability is then about 1 - exp(-0.023249416) = 0.0229812 or roughly 2.3%.

[Madpoo]

Originally Posted by MacPrime

To be honest, the one and only reason for my „upcoming announcement of M51“ was to make some fire under the GIMPS guys' asses. This seemed to pay off and speed up the process of M50. Only a few hours after my post M50 was finally announced.

Really? If you want to imagine you lit a fire to force the release on M50 to happen faster, and that makes you feel better, well... I won't burst your bubble. Must be nice in your world.

EDIT:
---
I'll add some info just for some background... a little 'behind the scenes' if you will. From the time the result was turned in on the 26th, it took me ~ 36-37 hours to do an unofficial verification starting at about 5 PM (pacific) -- unofficial because I was also using Prime95, like the original submission. My run finished at about 5:30 AM (pacific time) on the 28th. (all times below are also PST)

Ernst started the first of the official verifications with mlucas in the wee hours of the 27th and ATH started a run around 10 AM on the 27th (a cudalucas run that took around 72 hours and an mlucas run that, with a few stops/starts to modify the EC2 size I think took around 4.5 days (?)

Meanwhile, Mr. Pace sent his penultimate save file to George who re-ran the final iterations ... that was the first real sense we had that this was the real deal. That finished at 2:40 PM on the 27th.

My unofficial verification finished first, but we still had to wait for the official verification runs. AirSquirrels started a run with gpuOwL on the 28th at ~ 8 AM and since it was running on a super fast AMD FX Vega 64, it finished in ~ 35 hours and was the first of the official verifications. Since gpuOwL only has the 4M FFT size, there was a question on whether it would squeak through okay, but fortunately the interim residues from my run were matching his run all along the way.

AirSquirrels run finished at about 6 PM on the 29th, a run that ATH started using Prime95 (also unofficial) finished at 11:30 PM on the 29th, the ATH cudalucas run finished at 11 AM on the 30th, the ATH mlucas run finished at 2 PM on the 30th, Ernst's mlucas finished at 3:40 AM on the 31st.

In the meanwhile, there's the press release to work up, and as we've pointed out, releasing the news on New Years Eve means nobody would notice outside of our little circle, and that's no fun. There are also some server changes that have to happen (updating the milestone page, home page, known primes page, some DB changes that need to be done to "unmask" the new prime, etc.)

In short, I hope this illustrates that finding a new prime is a big deal, and there's a good amount of work being done behind the scenes to put all the parts in place, much of which falls on George's shoulders and it's the holidays, so give him a break guys! I mean, we didn't even have the "official" verification until later in the day on the 29th and into the 30th, and George is EST anyway... add 3 hours to all those times to get his perspective.

Oh, and one more thing... George had said, way back on the 28th, that he'd probably do the press release on Jan 3rd, for the reasons already explained. So, again, if you want to imagine you had anything to do with anything, keep smokin' the good stuff I guess. LOL

[ATH]

https://phys.org/news/2018-01-gimps-largest-prime.html

http://aperiodical.com/2018/01/27723...w-274207281-1/

https://www.newscientist.com/article...illion-digits/
(mostly paywalled)

https://www.reddit.com/r/math/commen...e_prime_found/

https://www.reddit.com/r/programming...me_discovered/

[MacPrime]

Thx, Aaron for the very detailed report of M50 behind the scenes.

I will try to be more patient with upcoming M51.

Keep up the good work!

[ixfd64]

The Guardian just published an article: https://theguardian.com/science/2018...han-23m-digits

[paulunderwood]

First YouTube video: .

Not much action, hard to read fonts, but a groovy soundtrack. I hope we get some from Numberphile...

[Siegmund]

Random question of the day:

When did/will the definition of "world record sized primes" in the assignment requester change?

Yes, I had a look over the holiday break at the size of assignments my machines set to world-record-sized were getting, to see if there was a big jump (or if I had been given an already-known-not-to-be-world-record-sized assignment I could gripe about.) No info, however: all of my assignments were in the 77.4-77.8M range both before and after the new prime came in.

[petrw1]

Originally Posted by paulunderwood

First YouTube video: .

Not much action, hard to read fonts, but a groovy soundtrack. I hope we get some from Numberphile...

They didn't notice that the "^" was dropped:
So 2^77,xxx,xxx-1 or even 2^5-1 was displayed as 277,xxx,xxx-1 and 25-1; and several others

[Prime95]

Originally Posted by Siegmund

Random question of the day:

When did/will the definition of "world record sized primes" in the assignment requester change?.

It changed December 30 -- give or take a day.