Holy new Mersenne prime, Batman! (M47 related)

[petrw1]

I noticed the Users counts on the stats page is up about 1,000 in the last couple days. Is that normal fluctuation or has there been a significant wave of new members since the media attention?

[Xyzzy]

I noticed the Users counts on the stats page is up about 1,000 in the last couple days. Is that normal fluctuation or has there been a significant wave of new members since the media attention?

We have no idea about new participants, but here is a graph of the forum's activity, measured in page hits.

[Prime95]

http://www.signonsandiego.com/news/s...1n18prime.html

[retina]

Originally Posted by http://www.signonsandiego.com/news/science/20080918-9999-1n18prime.html

2^4311269-1

Oops

[davieddy]

Originally Posted by retina

Oops

"They are the 45th and 46th primes discovered by GIMPS"

[Batalov]

I've had my small negative experience with S.D.Tribune a few years ago with the mouse genome... I hoped for the best this time... It's a good paper, just a little bit fast with facts. (as in 'pull a fast one' and 'scientific facts'). Ah well, those pesky digits are hard to get right. People need a good story, who cares about exact digits. We do!

On a bright side - at least it's not 243211269-1.

[ixfd64]

I just noticed something rather curious about M46. When the server is synchronized twice a month, any results that are older than two weeks are removed.

If you check the cleared.txt page, you will see that almost all of the results were returned in September. M46 (which was discovered in August) was retained, probably due to the verification process.

However, one other result from August was also kept.

38859463 69 0x5840B65CA7CB7B__ 23-Aug-08 07:23 curtisc JCKL-cce41L

Incidentally, it was returned ten minutes before M46 was reported.

[Mini-Geek]

Originally Posted by ixfd64

I just noticed something rather curious about M46. When the server is synchronized twice a month, any results that are older than two weeks are removed.

If you check the cleared.txt page, you will see that almost all of the results were returned in September. M46 (which was discovered in August) was retained, probably due to the verification process.

However, one other result from August was also kept.



Incidentally, it was returned ten minutes before M46 was reported.

I think George said that he put a fake residue on the number you quoted there, and set its status to prime (something about the prime count IIRC).

[jinydu]

Originally Posted by petrw1

I noticed the Users counts on the stats page is up about 1,000 in the last couple days. Is that normal fluctuation or has there been a significant wave of new members since the media attention?

Unfortunately, most news articles don't seem to emphasize that GIMPS is a distributed computing project that relies on volunteers. As a result, most people who hear the news don't get the message that anyone can join.

On another note, many news articles are also making the mistake of calling the verification computers "supercomputers". While certainly more powerful than your typical desktop or laptop, I think the word supercomputer is a big exaggeration.

[Jeff Gilchrist]

Originally Posted by jinydu

Unfortunately, most news articles don't seem to emphasize that GIMPS is a distributed computing project that relies on volunteers. As a result, most people who hear the news don't get the message that anyone can join.

Yes, the journalists seem to be leaving that part of the story out. There should be another article out on Saturday or Monday that goes more into the distributed computing and Internet community around GIMPS in the Globe and Mail. The journalist was supposed to be talking to George and Scott, not sure if he got a hold of Edson or not.

Originally Posted by jinydu

On another note, many news articles are also making the mistake of calling the verification computers "supercomputers". While certainly more powerful than your typical desktop or laptop, I think the word supercomputer is a big exaggeration.

I guess that depends on what you consider a "supercomputer". The TOP100 list systems have thousands of processors. Most non-technical people would probably think of the verification machines being "supercomputer"-like compared to their home systems. The verification machine I used has 128CPUs which is significantly more than a home machine, but not at the level of the TOP100 so where do you draw the line? I think "supercomputer" is a buzz word that people can relate to and journalists like to use as it is easier to understand even if it not entirely accurate than, "Computer that is more powerful than your home computer, but not quite a super computer".

[Housemouse]

I define a "supercomputer" as any computer that makes the semiannual list of the top 500.