P.I.E.S. - Prime Internet Eisenstein Search

[Batalov]

Holy mother of god, yes.
Thanks!

Next stop -- Phi(3,-b^524288) ...and before that perhaps Phi(3,-b^393216)

[paulunderwood]

Originally Posted by Batalov

Next stop -- Phi(3,-b^524288) ...and before that perhaps Phi(3,-b^393216)

Around 6 million digits?

[paulunderwood]

Another one with over 3 million digits. Congrats.



The previous one has yet to be UTM-verified.

[pepi37]

Originally Posted by paulunderwood

Another one with over 3 million digits. Congrats.



The previous one has yet to be UTM-verified.

When Batalov finished this work, we dont have any prime to find :) I am not familiar of what kind of monster ( read computer) can do such computation: last one is verified over three days, and it is just one candidate... and now we got new one in matters of days.
That computer power is something it is totally unreachable for us, mortals :)

[axn]

Originally Posted by pepi37

When Batalov finished this work, we dont have any prime to find :) I am not familiar of what kind of monster ( read computer) can do such computation: last one is verified over three days, and it is just one candidate... and now we got new one in matters of days.
That computer power is something it is totally unreachable for us, mortals :)

I believe the actual computation is being performed by a modified P95 (using multihtreading?) or a modified LLR, that is much faster than PFGW (which is being used by T5K site). But the real "secret" is the huge amount of compute power at the disposal of ryanp. I hear that it is bigger than NSA

[pepi37]

Originally Posted by axn

I believe the actual computation is being performed by a modified P95 (using multihtreading?) or a modified LLR, that is much faster than PFGW (which is being used by T5K site). But the real "secret" is the huge amount of compute power at the disposal of ryanp. I hear that it is bigger than NSA

Even that is 500% or 1000 % times faster then PFGW it is still for us mortals, unreachable :) Someone mush push little red button ( and below that button says)- Before you press it please, include a new nuclear power plant just for this purpose

Ryanp- can you borrow me your computer : I promise I will be finished in two hours

[Batalov]

I feel a disturbance in the force.

[paulunderwood]

Originally Posted by Batalov

I feel a disturbance in the force.

[axn]

Originally Posted by Batalov

I feel a disturbance in the force.

How many midichlorians?

[Batalov]

TOP12th# llr -d in >& out &
TOP12th# tail -f out
Base factorized as : 2^2*...*...
Base prime factor(s) taken : 2, ...
Starting N-1 prime test of 1NNNNN^786432-1NNNNN^393216+1
Using generic reduction FMA3 FFT length 1440K, Pass1=320, Pass2=4608, a = 3
1NNNNN^786432-1NNNNN^393216+1, bit: 10000 / 13XXXXXX [0.0Y%].  Time per bit: 42.735 ms.
...

Looks like a 4-5 day verification run...
Perhaps it's time to hack the multithreaded llr binary.

P.S. Cyclo is not written for this form: 393216 is not a power of 2, but 3 times a power of 2

[paulunderwood]

? 13000000/log(10)*log(2)
3913389.9436317555377786056314184093480
? 14000000/log(10)*log(2)
4214419.9392957367329923445261429023747

? #digits((100000)^786432)
3932161
? #digits((200000)^786432)
4168900

Any better sleuths out there?