Report top-5000 primes for k=1003-3000

[mdettweiler]

Originally Posted by gd_barnes

The chances of a prime in port 8000 at the current level are around 1 in 6700 so that is quite normal. I had a recent stretch where I found 1 prime in about 35000-40000 tests on that port. Right after I pulled my machines off of it, a huge run of primes came last night with at least 2 (maybe 3) new searchers finding one. I joked that I cleared out all of the primeless area for others.

I then switched to port 5000 and found a prime within a few hundred tests with a chance at worse than 1 in 7000. As MrOzzy found, it all evens out in the long run.

For the mini-drive, the chances are going to be around 1 in 5000-6000 tests.


Gary

Regarding everything evening out in the long run: ah, if only it would do that for me! It seems that I'm hitting all the primeless areas and everyone else is picking the primes out from in between them, regardless of what drive I'm running.

For example, my quad can produce approximately 1000 tests per day at port 8000's current level. That comes out to, on average, one prime in 6.7 days--round that up to 7 to account for the fact that my quad produces a wee bit less than 1000 tests (more like 930-950) per day. I've had it on port 8000 since close to the beginning of this month--more than 3 weeks of constant testing on that port. Yet, I've only pulled out one prime so far! I know these things even out over the long run, but then listen to this: Before that, I had the same quad on IB4000 (which Gary quoted the odds for as "worse than 1 in 7000") for more than 2 months. Based on what I remember being able to crank out on that port, I calculated that (assuming a 1 in 7100 odds of prime) I *should* be able to pop out one prime every 10 days. However, in all of those ~2.5 months I only found one!

I guess I'm long overdue for a big run of primes by now--hey, these things have got to balance out sometime or other!

Max

[Razor_FX_II]

Sorry about that gd_barnes. I knew it wasn't for the top 5000 and I new is fit in the 'k' value for this thread but didn't notice that it did not fit the 'n' requirement for this thread.
My apologies, working on 3 hours sleep in the last 48 hours isn't helping me any.

[gd_barnes]

Originally Posted by Razor_FX_II

Sorry about that gd_barnes. I knew it wasn't for the top 5000 and I new is fit in the 'k' value for this thread but didn't notice that it did not fit the 'n' requirement for this thread.
My apologies, working on 3 hours sleep in the last 48 hours isn't helping me any.

No big deal. You should ask David (IronBits) about his trials and tribulations early in the project trying to figure out where and what primes to show in what threads. He even joked at one point something like: "You just need a thread called David's primes" and "I'll dump them all there". Looking back on it, it was rather comical although I'm sure a little annoying to him at the time.

[MyDogBuster]

Sorry about that gd_barnes. I knew it wasn't for the top 5000 and I new is fit in the 'k' value for this thread but didn't notice that it did not fit the 'n' requirement for this thread.
My apologies, working on 3 hours sleep in the last 48 hours isn't helping me any.

Look at it this way. It keeps me in work. LOL

I had the same problems as David did when I first started. It took me a week to figure out the difference between a k and an n.

[PCZ]

1909*2^560435-1 New Prime

[IronBits]

Added 87843 : 1627*2^563789-1 (169721 digits)

[pfb]

1893*2^565191-1 is Prime! It appears to be 170143 digits long

(hope this is the right thread :p)

[MyDogBuster]

1893*2^565191-1 is Prime! It appears to be 170143 digits long

(hope this is the right thread :p)

Right thread. Nice one.

[IronBits]

Added 87848 : 1867*2^565489-1 (170233 digits)

[vaughan]

1665*2^567736-1 prime already in database.

[IronBits]

Added 87860 : 1735*2^568151-1 (171034 digits)