Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummerslowrise

11)
Message boards :
Number crunching :
Badges III
(Message 144679)
Posted 232 days ago by Gelly

12)
Message boards :
Number crunching :
WW testing
(Message 144389)
Posted 237 days ago by Gelly
Seems like robish was rearing to go on some NvidiaX test, so I'll also reserve CPU multithreading for Linux (with a 3970x, no less!)
Done!
CRC 1) 67029FBFBC1AEFB0
CRC 2) 2F32CAD23BA5B1CA
CRC 3) B8B040FA3813FEB2
Same thing as my Nvidia results. Checkpointed a lot more often, though, and failed to give an ETA at checkpoints like Nvidia did. Likely more down to implementation than any bugs.
The 3970X at 64 threads kept up with or outperformed the GTX 960 in terms of time, which is probably the wildest thing I've seen for CPU vs. GPU. Probably due to the fact that, once you get past 16 cores, the CPU gets closer to being a GPU than anything else.

13)
Message boards :
Number crunching :
WW testing
(Message 144387)
Posted 237 days ago by Gelly
Seems like robish was rearing to go on some NvidiaX test, so I'll also reserve CPU multithreading for Linux (with a 3970x, no less!)

14)
Message boards :
Number crunching :
WW testing
(Message 144386)
Posted 237 days ago by Gelly
I can reserve NvidiaLinux
Done!
CRC 1) 67029FBFBC1AEFB0
CRC 2) 2F32CAD23BA5B1CA
CRC 3) B8B040FA3813FEB2
The special instances expected in the document seemed to line up pretty well (test 2 also nabbed the two wieferich primes).
Done with a GTX 960 in a decent amount of time.

15)
Message boards :
Number crunching :
WW testing
(Message 144383)
Posted 237 days ago by Gelly
I can reserve NvidiaLinux

16)
Message boards :
Number crunching :
Évariste Galois's Birthday Challenge
(Message 144064)
Posted 247 days ago by Gelly
Any bets how many DIV primes we will find during the five days? /JeppeSN
I will invest an absurd amount of my personal time and money to get as close to ensuring at least one prime as I can.

17)
Message boards :
General discussion :
Favourite Prime Number.
(Message 143078)
Posted 281 days ago by Gelly
https://factordb.com/index.php?id=1100000001571209570 points to the current entry (and eventual proof) that 68686868686868....{4131 68's}67 is a prime number. Took a 7 hour bite out of my current Primo project, which is relatively nothing at all.

18)
Message boards :
General discussion :
Favourite Prime Number.
(Message 143025)
Posted 282 days ago by Gelly
(10^(2*4132)1)*68/99  1 has 4131 68's in a row, followed by 67. It's about 8k digits and hasn't been proven, but I'm waiting until pfgw finishes the range I gave it before I throw primo at the biggest one.

19)
Message boards :
General discussion :
Leyland primes
(Message 142605)
Posted 297 days ago by Gelly
Is there a possibility doing ECPP via distributed computing? I only found "Also ECPP has never been truly exposed to distributed computing" in this mersenneforum thread.
I'm not an expert on ECPP, but from the work I've done and what I've seen, it's definitely a possibility. In 2005, they were using multiple computers in tandem for fastECPP, although definitely a far cry from distributed computing.
The way the main stage of ECPP works is it essentially connects the large prime to a small prime through a bunch of intermediate primes with elliptic curves that have special properties. It's essentially a random walk, hoping to hit on an equation that provides a factorization that you can work with, and many times Primo will end up backtracking if it doesn't like the current prime it's working on. This stage would likely be difficult to distribute well, but with enough distribution, maybe "ok" is good enough.
The second stage is a lot easier to split up and lends itself much better to distribution, as it's just taking all the equations to connect the primes in stage one and factorizing them, AFAIK.
Again, I'm no expert on ECPP, either mathematically or programmatically, and considering the maelstrom of ECC patents and Primo being closed source, it's hard to blame me. Perhaps we could change that for the better if we got enough people booked up on elliptic curves looking into distributed computing  it may be possible to blow past the current limits of ECPP due to it being single system.

20)
Message boards :
General discussion :
Leyland primes
(Message 142565)
Posted 300 days ago by Gelly
I generally don't like the idea of new subprojects, but if ever someone feels the absolute need to implement one, Leyland primes have a certain appeal.
They are of the form x^y + y^x and the biggest currently know one 8656^2929 + 2929^8656 has a mere 30008 digits.
According to PRP Records there are probable Leyland primes with up to 10^6 digits, but apparently it's hard to prove they are prime.
It might not be be easy to make it work with DC, the test for each number would have to be broken down into tasks. Maybe it doesn't make sense if they cannot be run parallel. But it irks me that 99% of new large primes are of a specific form just because those are relatively easy to prove their primality.
It's definitely out of character for PrimeGrid to pick up a subproject that cannot prove the primes they are testing, as all of the current and past projects are either sieves or for provable primes.
The reason Leyland primes aren't helpful in terms of primality proving is that you don't tend to know a lot about the factorization of P1 or P+1, which is used in many of the primality proofs for very large primes. You can use general primality proving techniques, such as ECPP, but those top out at about 40k digits at the moment (although there are stirrings of 50k digits in the next 5 years)

Next 10 posts
