Oops - New Prime! (M49 related)

[petrw1]

Originally Posted by Prime95

While you are waiting, as a reward to GIMPS' most dedicated users, a hint:

The second digit is a 4

And I bet the last digit is either: 1, 3, 7, 9

[firejuggler]

so we know it is in the 74M

[ewmayer]

Originally Posted by Prime95

While you are waiting, as a reward to GIMPS' most dedicated users, a hint:

The second digit is a 4

Echoing Wayne's reply to your post - 2nd digit of the exponent or of the number's decimal expansion?

[lavalamp]

Originally Posted by petrw1

And I bet the last digit is either: 1, 3, 7, 9

So on average the last digit is 5.

The rest of the dominoes will fall like a house of cards. Checkmate.

[CuriousKit]

Heh, well, even if it is the exponent, it could be in the 64M range. We have been clearing that lot as of late, so who knows.

Either way, congratulations to whoever found it first and to those who went to verify it.

I kind of see the LL tests like buying a lottery ticket... you don't except to score big and hit a prime, but it's nice if you do. And unlike buying a lottery ticket, you are putting work and effort towards a global cause that most people can acquire eat least a passing understanding of.

[firejuggler]

see' http://www.mersenneforum.org/showpos...&postcount=103

[Batalov]

Originally Posted by ewmayer

Echoing Wayne's reply to your post - 2nd digit of the exponent or of the number's decimal expansion?

The decimal expansion contains the Answer to the Ultimate Question of Life, the Universe, and Everything (42)

[Madpoo]

Originally Posted by ATH

More nostalgia while we wait. Here people in 2002/2003 are talking big about how fast their computers are now on the 16M-20M exponents compared to GIMPS beginning in 1996-1998...

That reminds me that I ran across the story of the Cray finding M32 ( M(756839) ) in "only" 19 hours on a Cray 2, and a mere 3 hours on a Cray C90 with 16 CPUs.
https://primes.utm.edu/notes/756839.html

So I was curious... how fast was it on a modern CPU. Well, I finished in 3 minutes give or take a few seconds.

So... that compares a super fast Cray of 1992 vintage to today's super fast servers, and I'm sure even a decent desktop could do it in < 5 minutes.

And you're right... in a similar time frame (14 years from now) people will look at the 34 hours it took me to verify M49 and be amazed that it only took 34 seconds. Their 16K core MIC chip with memory bandwidth measured in hundreds of TB/second will smoke anything we have now.

[axn]

Originally Posted by Prime95

While you are waiting, as a reward to GIMPS' most dedicated users, a hint:

The second digit is a 4

Originally Posted by Batalov

Actually, the same can be said about another WR search effort... Hmmm...
In it, b needs to be > SIKRIT_b_value to be a contender again.

So SIKRIT b value > 200000 but < 250000

[Mark Rose]

Originally Posted by Madpoo

That reminds me that I ran across the story of the Cray finding M32 ( M(756839) ) in "only" 19 hours on a Cray 2, and a mere 3 hours on a Cray C90 with 16 CPUs.
https://primes.utm.edu/notes/756839.html

So I was curious... how fast was it on a modern CPU. Well, I finished in 3 minutes give or take a few seconds.

So... that compares a super fast Cray of 1992 vintage to today's super fast servers, and I'm sure even a decent desktop could do it in < 5 minutes.

And you're right... in a similar time frame (14 years from now) people will look at the 34 hours it took me to verify M49 and be amazed that it only took 34 seconds. Their 16K core MIC chip with memory bandwidth measured in hundreds of TB/second will smoke anything we have now.

My rule of thumb is that a super computer is on your desk in 10 years, your lap in 20, and your pocket in 30.

[E_tron]

I eagerly await the news. Congrats to the Mersenne Community; a genuine discovery.

Too bad we have to wait. If it was a hash, anyone could verify it . It's unfortunate so many resources are spent chasing vanity 0x000000... hashes in exchange for some digital coupons.