Riesel base 6 - team drive #4 - EIGHT OR BUST!

[Batalov]

k=1597 complete to 1M, no primes.

[Batalov]

Taking 521-530K (with 4 k's).

[Batalov]

Guys?
This number was waiting for three months for anyone who could and should have taken this range:

51017*6^528803-1 is 3-PRP! (4189.4352s+0.0405s)


Verifying now.
Where is the dancing banana?

[mdettweiler]

Originally Posted by Batalov

Guys?
This number was waiting for three months for anyone who could and should have taken this range:

51017*6^528803-1 is 3-PRP! (4189.4352s+0.0405s)


Verifying now.
Where is the dancing banana?

Sweet! Only 4 more to go.

[gd_barnes]

Originally Posted by Batalov

Guys?
This number was waiting for three months for anyone who could and should have taken this range:

51017*6^528803-1 is 3-PRP! (4189.4352s+0.0405s)


Verifying now.
Where is the dancing banana?

OMG...Way to rub it in. Meanie. And I thought about grabbing a range several times too. Dancing georges are my favorite so here goes:






Congrats! Wow, we're clearly above the expected # of primes now! We're far enough ahead of it that I think your PRP may be composite. lmao

Actually, I think there would not have been a prime if I had tested it. n>1M base 2 (300K+ digits) has been very unkind to me...haven't gotten one yet. I haven't tried a huge # of tests but enough to be annoyed by not getting one yet.

Now, I'm curious to see how long it takes to prove at the top-5000 site. I've suggested to Prof. Caldwell that they need to upgrade to newer versions of PFGW for proof but they're still using older versions. It'll likely be well over a day.

Congrats again!


Gary

[MyDogBuster]

51017*6^528803-1 is 3-PRP! (4189.4352s+0.0405s)

Nice one Serge

[Batalov]

...it is only by standing on the shoulders of giants, you know. Thank you.

It also checks out as a 7-PRP (on another computer, to be sure, and with another FFT size), so I submitted it.
-tp will finish tomorrow and UTM's independent check sometime later.

[gd_barnes]

Originally Posted by Batalov

...it is only by standing on the shoulders of giants, you know. Thank you.

It also checks out as a 7-PRP (on another computer, to be sure, and with another FFT size), so I submitted it.
-tp will finish tomorrow and UTM's independent check sometime later.

At this moment, at 411494 digits, it will come in at 155th place with a score of 25! You nearly just doubled your entire personal score. Wee!

[mdettweiler]

Originally Posted by gd_barnes

Congrats! Wow, we're clearly above the expected # of primes now! We're far enough ahead of it that I think your PRP may be composite. lmao

I've noticed that, consistently, conjecture searches like this seem to do rather better than the expected # of primes. Five or Bust is another good example of this: they're all the way down to their final k yet are still 10s of M's behind the n-level where they would have expected to be for that. I have to wonder if there's more to this than the simple random distribution that we're assuming in our projections.

[gd_barnes]

Originally Posted by mdettweiler

I've noticed that, consistently, conjecture searches like this seem to do rather better than the expected # of primes. Five or Bust is another good example of this: they're all the way down to their final k yet are still 10s of M's behind the n-level where they would have expected to be for that. I have to wonder if there's more to this than the simple random distribution that we're assuming in our projections.

Funny. My thinking goes the other way. Those are two extreme exceptions from my experience. Look at all of the bases here that have one k remaining where that k has been searched to an n-level 10 to 100 times or more than the largest prime found on the base so far.

My feeling if anything...in the long run, you should expect somewhat less than the expected # of primes on low weight k's. That's because IMHO, if there are more small factors (which causes their low weight), then there's a better chance that there are more large factors (i.e. above a normal sieve limit). I can't prove it though. That's part of what NPLB is all about...attempting to figure out if high-weight k's yield more prime per candidate tested after sieving to a specific depth.


Gary

[Mini-Geek]

Originally Posted by gd_barnes

Funny. My thinking goes the other way. Those are two extreme exceptions from my experience. Look at all of the bases here that have one k remaining where that k has been searched to an n-level 10 to 100 times or more than the largest prime found on the base so far.

Perhaps it's not a simple 'yes or no', but based on some other factors that we don't know to look for yet.

Originally Posted by gd_barnes

My feeling if anything...in the long run, you should expect somewhat less than the expected # of primes on low weight k's. That's because IMHO, if there are more small factors (which causes their low weight), then there's a better chance that there are more large factors (i.e. above a normal sieve limit). I can't prove it though. That's part of what NPLB is all about...attempting to figure out if high-weight k's yield more prime per candidate tested after sieving to a specific depth.


Gary

To that end, has someone done that sort of comparison yet? To see how many primes there are per unsieved candidate vs how many we think we should expect?