[quote=T.Rex;177117]I have Glucas running on Nehalem 8 cores.
Verification result should be available Friday 1:45 pm USA CST.
Tony[/quote]

I hope you have activated 16 threads for it. I don't believe it will be Friday - Prime95 will probably estimate the enddate wrong because of the hyperthreading.

[quote=jinydu;177153]What does GA stand for?[/quote]
I'm thinking General Availability (i.e. he doesn't want to give out information about a CPU that's not released yet for some reason, I'd guess related to it possibly being a pre-final CPU and perhaps some sort of NDA). It's the only "GA" I can think of that would b applicable.
[quote=joblack;177158]I hope you have activated 16 threads for it. I don't believe it will be Friday - Prime95 will probably estimate the enddate wrong because of the hyperthreading.[/quote]
He's using Glucas and is probably calculating the end date himself based on the timings from Glucas.

Glucas/Nehalem : 75% done - 32M
 

[QUOTE=Mini-Geek;177160]I'm thinking General Availability ... NDA.
He's using Glucas and is probably calculating the end date himself based on the timings from Glucas.[/QUOTE]Yes !

Nice guesses !

[QUOTE]Saved Interim file at iteration 32000000. Res64: 7844B88D659F9B61.
[/QUOTE]

Have a good night !

Tony

I think someone brought this up earlier (perhaps in a different thread) but the program (Prime95, Mlucas, Glucas, etc) can check the whole residue, right? Someone indicated there is a small chance in which the 64 bit residue at the end could have 16 zeroes but yet the residue prior to that would have alpha-numeric characters. However, they also said that the chances of this were nearly infinitesimal. I just wondered if there was a safe-guard built in to recognize this (very rare) scenario in case in happened.

[quote=Mini-Geek;177148]
[URL="http://www50.wolframalpha.com/input/?i=2009-06-12+1%3A45+pm"]T-24:08:37[/URL]
:bounce:
[/quote]

Don't want to dampen anyone's sense of anticipation,
but when Tony finishes, he will tell us he has verified that
the number he was testing is prime.
George will probably wait for the second verification before
announcing the exponent -and then only if another prime isn't
discovered before, when he will wait for that to be verified too:smile:.

[quote=Primeinator;177165]I think someone brought this up earlier (perhaps in a different thread) but the program (Prime95, Mlucas, Glucas, etc) can check the whole residue, right? Someone indicated there is a small chance in which the 64 bit residue at the end could have 16 zeroes but yet the residue prior to that would have alpha-numeric characters. However, they also said that the chances of this were nearly infinitesimal. I just wondered if there was a safe-guard built in to recognize this (very rare) scenario in case in happened.[/quote]
Yes, the program (at least Prime95) can check the whole residue, and yes Prime95 does before reporting a number as prime. And the chance is around (exactly, I think, not too certain) 2^64 to 1, so I'd call that infinitesimal.

[QUOTE=Primeinator;177165]I think someone brought this up earlier (perhaps in a different thread) but the program (Prime95, Mlucas, Glucas, etc) can check the whole residue, right? Someone indicated there is a small chance in which the 64 bit residue at the end could have 16 zeroes but yet the residue prior to that would have alpha-numeric characters. However, they also said that the chances of this were nearly infinitesimal. I just wondered if there was a safe-guard built in to recognize this (very rare) scenario in case in happened.[/QUOTE]
I can only speak for myself, but Mlucas only reports "prime" if the entire residue R = 0; if the residue were nonzero but the bottom 64 bits were = 0, it would report "Mxxxxxxxxx is not prime. Res64: 0000000000000000."

I also have the code report the Selfridge-Hurwitz residues traditionally used for Fermat-number Pe'pin test results, defined as R%(2^35-1) and R%(2^36-1); thus the odds of all 3 reported "short" residues being 0 but the number not being prime drop to roughly 1 in 2^135.

[QUOTE=Primeinator;177165]I think someone brought this up earlier (perhaps in a different thread) but the program (Prime95, Mlucas, Glucas, etc) can check the whole residue, right? Someone indicated there is a small chance in which the 64 bit residue at the end could have 16 zeroes but yet the residue prior to that would have alpha-numeric characters. However, they also said that the chances of this were nearly infinitesimal. I just wondered if there was a safe-guard built in to recognize this (very rare) scenario in case in happened.[/QUOTE]

1 in 2[sup]64[/sup] or 1 in 18 billion billion (assuming each 64bit residue has equal chance which they probably don't). All programs has the entire residue offcourse or they wouldn't be able to continue LL test. If you saw 0x0000000000000000 as 64-bit residue and the program continued it would be just the last 64+ bits that was zero. If the whole p-bit residue was zero, the test would go into the loop 0, -2, 2, 2, 2, 2, 2 ..... and programs would probably give an error. But I doubt its possible Sq = 0 (mod 2^p-1) with q<p-2.

[QUOTE]1 in 264 or 1 in 18 billion billion (assuming each 64bit residue has equal chance which they probably don't). All programs has the entire residue offcourse or they wouldn't be able to continue LL test. If you saw 0x0000000000000000 as 64-bit residue and the program continued it would be just the last 64+ bits that was zero. If the whole p-bit residue was zero, the test would go into the loop 0, -2, 2, 2, 2, 2, 2 ..... and programs would probably give an error. But I doubt its possible Sq = 0 (mod 2^p-1) with q<p-2.[/QUOTE]

Okay, thanks. I kind of figured as much but I just wanted to be sure...I can kind of be OCD about certain things at times. And I should have known the program could check the whole residue... as you pointed out it could not continue the LL without this. Sometimes I misplace my brain too!

I see. T.Reix is probably using Beckton:

[url]http://en.wikipedia.org/wiki/Nehalem_(microarchitecture[/url])

[quote=jinydu;177193]I see. T.Reix is probably using Beckton:

[URL="http://en.wikipedia.org/wiki/Nehalem_%28microarchitecture"]http://en.wikipedia.org/wiki/Nehalem_(microarchitecture[/URL])[/quote]

Yeah, also known as Nehalem EX. Still I don't think that an 8-core cpu will take it in 2 days. My 4-core Q6600 needs about 10 days ... and it is overclocked.