LLRnet/PRPnet rally January 3rd-10th

[gd_barnes]

Originally Posted by mdettweiler

Indeed--just a 185-test difference between Lennart and Brian. Two additional cores could cover that difference in just a day.

With a 110-pair difference, they both just keep adding more and are now up to 380+/hour each now. This is the most interesting individual battle we've had in any rally so far.

[mdettweiler]

Another interesting stat: in two and a half days of solid processing on port 9000, we have not found a single prime. As of the DB's last hourly update, we've tested 30,130 candidates on port 9000 during the rally. According to Gary's odds-of-prime spreadsheet, there is a 95.188% chance of finding one or more primes in that many candidates at n=890K and p=1405T sieve depth, with 3.034 primes expected. Talk about underperforming! Hopefully we'll grab a nice big bunch of primes soon and even things up a bit. (Maybe I'll finally get my long-hoped-for rally prime! )

BTW @Gary: were we going to do 24-hour LLRnet vs. PRPnet mini-contests throughout the rally like we did last time?

[gd_barnes]

Originally Posted by mdettweiler

BTW @Gary: were we going to do 24-hour LLRnet vs. PRPnet mini-contests throughout the rally like we did last time?

No for obvious reasons.

[mdettweiler]

Originally Posted by gd_barnes

No for obvious reasons.

Ah, right...duh, silly me.

[Steve]

Can not say that any more

687*2^931420-1 is a prime on port:9000 @ Thu Jan 6 16:58:08 Digits-280,388

Is this a top 5K?

Do I need to report or will you?

Steve Martin

[Mini-Geek]

Originally Posted by Steve

Can not say that any more

687*2^931420-1 is a prime on port:9000 @ Thu Jan 6 16:58:08 Digits-280,388

Congrats!

Originally Posted by Steve

Is this a top 5K?

Yes, definitely! It will be #629 (unless any larger ones enter the list between now and when you report it and it's verified).

Originally Posted by Steve

Do I need to report or will you?

You should report it. If you haven't done it before, here are instructions.

[Steve]

687*2^931420-1 is a prime => confirmed
http://primes.utm.edu/primes/page.php?id=97279

[mdettweiler]

Originally Posted by Steve

687*2^931420-1 is a prime => confirmed
http://primes.utm.edu/primes/page.php?id=97279

Just to clarify a little bit of terminology: when we refer to a "confirmed" prime here we mean something slightly different than when the top-5000 site refers to a prime being "confirmed". Due to sporadic searches having been done over bits and pieces of this drive's range earlier on (before NPLB was founded), there are a few previously-known primes scattered throughout the range. We're re-doing the whole thing since quite often there were errors in these earlier searches; the upshot is that most of the primes we find are new, but every once in a while we run into one that was already known, in which case we just mark it as "confirmed".

As it turns out, Gary just cleaned up the last of the already-known primes in the 6th Drive a short while ago--you'll see it marked as "confirmed" in the list in the first post of this thread. So all further primes found in this rally (or even in the 6th Drive after the rally is done) will be new primes.

In the case of your prime (687*2^931420-1), you found a brand new prime. (If it was previously known, the top-5000 would have given you an "already in database" error.) Congratulations!

[AMDave]

Getting "No available candidates" on GB-9000
Is it log-jammed or out of work?

/ed -
scratch that. Just getting it on the 4 clients on one host: dn08c1, dn08c2, dn08c3, dn08c4
it connects
but it is giving me the no available candidates message
all the other clients on the other hosts are working
odd
if they all go dry, then I'll know it's GB-9000
- ed/

[AMDave]

got this from the debug log:
"[2011-01-08 18:25:28 EAST] 544: select indicates socket has data, but recv returned nothing"
Looks like GB-9000 is holding out on dn08.
That's 6 cores that are now doing something else. :(

[PCZ]

Dave check the permissions on the llrnet binary.
You can get that error if it isn't executable.