Carol / Kynea Coordinated Search - Reservations

[kar_bon]

I'm working on CK base 360-2000, n=1-1024 to get a point to start of (some smaller bases later filled). For now only the least n (n<=1024) for bases 1-1024 were evaluated and Batalov's search gave no list.

The PFGW-script looks like this:

ABC2 ($b^$a-1)^2-2 | ($b^$a+1)^2-2
a: from 1 to 1024
b: from 360 to 2000 step 2

My search for CK44 is currently at n=72k, continuing to 100k, new PRP (44^60212+1)^2-2 found so far.

[gd_barnes]

Originally Posted by LaurV

I was looking to those tables and saw that after 222 or so, the bases are not contiguous. I assume there is no search done for the missing bases, and the "higher" values just come from "people with a hobby" and not from an organized search. Because, unless I am not missing anything, I see no reason why the missing bases would not give primes. I think they do, and it should be no reason to jump from 222 to 228 (therefore missing 224 and 226) or from 2010 to 2026.

Now, the introduction done, for a week or so we were looking for an opportunity to test the new "multi-threaded" tool from Mark (i.e. mtsieve). We picked CK numbers, and we picked the base 2018. As the current year, you know? We could not make the toy run in multi-threaded mode (this is subject for another thread), but we got hooked and let it run overnight to sieve (single thread) and test with pfgw with some batch file. In the morning we had few (small) primes, and pfgw was close to testing n=12k or so.

Does it make any sense reporting it? Or we are really missing the point (again, it won't be the first time, hehe).

That is correct. If you do not see a base listed, it has not been tested. There are plenty of "base holes" to fill so feel free to reserve and test some if you would like.

I'm happy to hear that you guys are working on the multi-threaded version of cksieve.

Mark had previously asked that people only reserve and report bases if you intend to test them to n>=10K. I would like to stick with that requirement so as to not have a lot of admin work for small tests. So if you have a base that you have tested to n=12K, yes please report its results.

[gd_barnes]

Originally Posted by kar_bon

I'm working on CK base 360-2000, n=1-1024 to get a point to start of (some smaller bases later filled). For now only the least n (n<=1024) for bases 1-1024 were evaluated and Batalov's search gave no list.

The PFGW-script looks like this:

ABC2 ($b^$a-1)^2-2 | ($b^$a+1)^2-2
a: from 1 to 1024
b: from 360 to 2000 step 2

Do as you please but something like this has already been done twice before now; by both Serge and Sweety.

Serge's effort was done in 2016.

I feel like you are reinventing the wheel a 2nd time here. :-)

You'll need to reserve a specific base to n=10K for me to show it on the page.

[kar_bon]

I know of those post and as mentioned, both of them gave no list of all primes upto n=1024, only the first one to this bound or higher ones if none was found then.
They invented the wheel perhaps, but gave no complete instructions for a whole wheel.

[Dylan14]

Base 6 is at n = 124223. No new primes have been found yet, and I am continuing to n = 150k.

[gd_barnes]

Base 46 is complete to n=30K. No primes were found for n=10K-30K. Base released.

Base 46 was also doublechecked to n=10K. One missing prime was found and already reported.

Reserving base 48 to n=30K.

I will doublecheck it to n=10K.

[gd_barnes]

I have attached two available sieve files to the first post of this thread. Thanks to Dylan for pointing them out. We will use that method for the time being for making them easily accessible. If they become too numerous we will look to add a column to the main status page.

[sweety439]

Reserve 362, 364, 368, 394, 426 and 472 to n=10K.

[gd_barnes]

Base 48 is complete to n=30K. 4 primes were found for n=10K-30K. Base released.

Base 48 was also doublechecked to n=10K. No problems found.

All bases <= 50 are now complete to n>=30K. :-)

[gd_barnes]

Jiahao He has completed bases 206, 208, 210, and 212 to n=10K. Primes reported in other thread. He is releasing these bases.

All bases <= 256 are now complete to n>=10K.

[Dylan14]

I have attached here a sieve file for base 290 for the n range of 20k to 30k. It is sieved up to 2.1 T and should be ready to go for PRP/prime testing.
After I complete a reservation for factorizing a repdigit-related number, I will create some more sieve files. If possible Gary, could you tell me what bases and n-ranges are most wanted?