Bases 6-32 reservations/statuses/primes

[Xentar]

1 or 2 months ago, I started Sierpinski, Base 18.
Think, I will continue this search.

k = 122
Sieved till 4P (can anyone tell me, how to calculate the optimal sieve depth?)
LLR till n=128000

[gd_barnes]

Originally Posted by Xentar

1 or 2 months ago, I started Sierpinski, Base 18.
Think, I will continue this search.

k = 122
Sieved till 4P (can anyone tell me, how to calculate the optimal sieve depth?)
LLR till n=128000

Hi Xentar; welcome to the effort! I'll show this one as reserved by you.

Quick question...Have you already LLR'd until n=128000 or is that how far you are going to LLR it in the future?

You should sieve until the rate at which candidates are being removed approximately equals the rate at which they can be LLR'd.

To be very specific, you should do an LLR test of ONE candidate that is at 70% of the n-range that you are sieving and sieve until the removal rate equals that testing time. So if you are sieving n=100K-200K, you should run an LLR test on a candidate around n=170K (one that already has small factors removed). However long that takes to LLR is the removal rate you should sieve to. Sieving to P=4P (same as P=4000T if I remember correctly) sounds a bit high but I don't know what range of n you are sieving.

Any more questions...just ask. Everyone's quite helpful around here.


Gary

[Xentar]

Hi,

yes, I have already LLR'd to n = 130000, and I will try to do it, until I find a prime ;)

I sieved till n = 1,000,000. At the moment, I get a rate of 20 minutes per factor, and I need about half an hour per LLR test. So I think, I can continue sieving a while. Ok, thank you.

Edit:
Another question. When I started Sierpinski Base 18, I was told that there are three more k: 18, 324 and 381. For 381 I found a prime - but what is with 18 and 324? There were some "problems", because the k = base, but don't we need to find a prime here, too?

[gd_barnes]

Originally Posted by Xentar

Hi,

Another question. When I started Sierpinski Base 18, I was told that there are three more k: 18, 324 and 381. For 381 I found a prime - but what is with 18 and 324? There were some "problems", because the k = base, but don't we need to find a prime here, too?

For now, we are not searching k's that are multiples of the base due to some of the issues involved. In this case, k=18 and k=324 for Sierp Base 18 are Generalized fermat #'s and are unlikely to yield a prime with current technology. See the last para. of the project definition in the 'Come join us' thread. The same issue exists for Sierp Base 22 for k=22 and k=484.

We will take up the issue of multiples of the base after the project has been going for a month or two.


Thanks,
Gary

[Xentar]

Ah ok, thank you.
So I will just do k = 122

[mdettweiler]

Woo hoo! I found a prime! And it's my first prime ever!

I don't think it qualifies for top-5000 status, but here's what I found in my lresults.txt file today as a nice little Christmas present:

4885*2^243388+1 is prime!

This prime is equivalent to 4885*16^60847+1; this means that k=4885 has been eliminated for Sierpinski Base 16!

k=4885 has been searched up to n=60973; it went a little over since I didn't notice the prime until a few more candidates for that k had been tested. I am unreserving k=4885; the remaining sieved candidates (up to n=100,000) are attached to this post in case anyone wants to search this k further.

Also included in the archive attached to this post are the residuals for both k=2908 and k=4885 up to n=60973 (for k=4885) and n=61188 (for k=2908). I am still searching k=2908.

[gd_barnes]

Originally Posted by Anonymous

Woo hoo! I found a prime! And it's my first prime ever!

I don't think it qualifies for top-5000 status, but here's what I found in my lresults.txt file today as a nice little Christmas present:

4885*2^243388+1 is prime!

This prime is equivalent to 4885*16^60847+1; this means that k=4885 has been eliminated for Sierpinski Base 16!

k=4885 has been searched up to n=60973; it went a little over since I didn't notice the prime until a few more candidates for that k had been tested. I am unreserving k=4885; the remaining sieved candidates (up to n=100,000) are attached to this post in case anyone wants to search this k further.

Also included in the archive attached to this post are the residuals for both k=2908 and k=4885 up to n=60973 (for k=4885) and n=61188 (for k=2908). I am still searching k=2908.

Yeehaw! 55 k's to go on Sierp base 16. I'll remove it from the big sieve.


Gary

[gd_barnes]

Cut-and-paste from tcadigan's prime find on Sierp base 16. Not moved here because it would be prior to first post.

Originally Posted by tcadigan

7773*2^199784+1 is prime! Time: 213.884 sec.

i.e.
7773*16^49946+1
60145 digits

I'll try my 6663 to 100K next

[gd_barnes]

Status from Carlos (em99010pepe):

404*12^n+1 is complete to n=79.6K.

He decided not to test 288*13^n-1 so that one is now available. It had already been tested to n=62K and a sieved file is available in the sieving thread up to n=100K.


Gary

[tnerual]

32673*2^101924+1 is prime! Time: 51.669 sec.
or 32673*16^25481+1

[gd_barnes]

Originally Posted by tnerual

32673*2^101924+1 is prime! Time: 51.669 sec.
or 32673*16^25481+1

Great; another down and 54 k's to go for Sierp base 16. That was a quick one; just above our prior testing limit of n=25K.

Are you searching any more k's right now? If so, I'll show them reserved by you to keep any double-work from happening. Also, if you want a file sieved to P=400G on any k that you're interested in for Sierp base 16, I can send it to you ~10-11 PM today CST US (4-5 AM Thurs. GMT) after our big sieve is done to that point.

Thanks for searching.


Gary