45th Known Mersenne Prime Found?

The GIMPS (Great Internet Mersenne Prime Search) seems to have found a new prime number; the 45th known Mersenne Prime. The discovery was reported on August 23rd, by a GIMPS client computer. Two verification runs are currently underway which are estimated to be completed by the 12th and 16th of September.

The previous Mersenne Prime - found on September 4, 2006 - was more than 9.8 million digits long. If this new prime breaches the 10 million digit mark, the GIMPS will be able to claim the Electronic Frontier Foundation's $100 000 award.

Check here for updates.

That is good news!

There are several BOINC projects looking for prime numbers too.

Good to post an occasional message or two here!

It now seems that M46 could have been found even before M45 could be verified! That's just unbelievable, given the amount of crunching that goes into finding each one of these monsters.

See : http://www.mersenne.org/prime.htm and http://www.mersenneforum.org/showthread.php?t=10616.

In the latter forum, the user called Prime95 is, in fact, the famous George Woltman himself http://en.wikipedia.org/wiki/George_Woltman.

I've suddenly rediscovered my prime-hunting feet after many years away from GIMPS (and numerous other prime projects). Primegrid has allowed participation in both BOINC and prime crunching.

Warped.

Hi Warped...

The first verification for M45 has been completed and they expect the second verification to complete today!! They expect to have M46 verified by tomorrow. Its all so close.

I wonder if either of them will breach the 10 million digit mark...

Hi stackoverflow.

Now we have to wait until next week while they find a suitable press outlet! I would have thought they could use their own website.

I am certain one of them is in excess of 10M digits, based on all the discussions that have been taking place. George has been dropping subtle hints that this is the case as well.

Warped wrote:...while they find a suitable press outlet!

When you said that I actually thought you were kidding until I read it on their website...

The announcement has finally been made! Both are greater than 10m digits.

http://mersenne.org/prime.htm

It certainly took a while for the news to break.

Spent the last few hours just reading up about prime numbers and related content. Its a seriously dense topic. I was initially interested only in finding out what the applications of primes are but found much more information. For those who are interested, here are just two resources which I assure you will lead you to many others:

The Prime Pages, maintained by Professor Chris Caldwell at the University of Tennessee at Martin.

Applications of Prime Numbers, on wikipedia. I wouldn't swear by the accuracy of this information, but the same could be said for any information found online. Either way its a start.

Happy reading!

Yes, indeed, Stack. This is a subject which seems to have no end. The site of Prof Chris Caldwell is certainly a mine of information (and links to many other sites).

I must say, though, that hunting for large primes requires a great deal of patience as the gaps between them grows larger, the bigger the numbers get, and the amount of crunching required for each test gets more and more as well. I'm currently trying to crunch a number 2 to the power of 47,464,777, minus 1 to see if it's prime. It will take a further 11 months to crunch!

By the way, Time magazine has voted the finding of the 46th (known) Mersenne Prime as the 29th best invention of 2008. See Here.

11 months?!

Yes I'm sure anyone who's willing to wait that long for an outcome is extremely patient. IHMO one shouldn't try to crunch numbers that big unless one has the necessary hardware resources. For that reason I will be giving any such endeavours a wide berth.

By the way, for which project is this? Primegrid?

stackoverflow wrote:11 months?!

Yes I'm sure anyone who's willing to wait that long for an outcome is extremely patient. IHMO one shouldn't try to crunch numbers that big unless one has the necessary hardware resources. For that reason I will be giving any such endeavours a wide berth.

By the way, for which project is this? Primegrid?

Hi Stack

Apologies for overlooking your question. the Project is GIMPS (a non-BOINC project). In other words the hunt for Mersenne Primes.

Warped wrote:Hi Stack

Apologies for overlooking your question. the Project is GIMPS (a non-BOINC project). In other words the hunt for Mersenne Primes.

Hi Warped

Do you leave the machine on 24 hours a day?

stackoverflow wrote:Hi Warped

Do you leave the machine on 24 hours a day?

Yes, I do, but not on GIMPS. In fact, it runs GIMPS only about 2 hours a day.

Warped wrote:... it runs GIMPS only about 2 hours a day.

Hi Warped.

Might I ask why you do so? Surely the task would finish far under 11 months if you ran GIMPS all the time.

Hi Stack

Two reasons:
1. I have a dual boot on my computer (Windows 32-bit and Linux 64-bit). GIMPS is running in Windows.
2. GIMPS prime crunching (as opposed to finding factors) is extremely hard on the CPU. The software (Prime95) is the benchmark for stress testing. In particular, in summer and during the day while I'm at work, I try to leave the computer with less stressful work to do.