mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Conjectures 'R Us (https://www.mersenneforum.org/forumdisplay.php?f=81)
-   -   2018 project goals (https://www.mersenneforum.org/showthread.php?t=22885)

gd_barnes 2018-01-04 10:59

2018 project goals
 
Below are some goals for 2018 that CRUS could shoot for:

1. Prove 10 bases;[COLOR=green] 8 bases to go.[/COLOR]

2. Search all bases <=700 with 1 k remaining (at n=200K) to n>=400K;[COLOR=green] 10 bases to go.[/COLOR]

3. Search all bases with <= 3 k's remaining (at n=100K) to n>=200K;[COLOR=green][COLOR=#ff0000]completed 2018/12/01[/COLOR][/COLOR]

4. Search all bases with < 400 k's remaining (at n=25K) to n>=100K;[COLOR=green][COLOR=#ff0000] completed 2018/08/29[/COLOR][/COLOR]

5. Search all bases with < 1000 k's remaining (at n=10K) to n>=25K; [COLOR=#ff0000]completed 2018/07/07[/COLOR]

6. Complete searching of S3 to n>=25K; [COLOR=red]completed 2018/02/24.[/COLOR]

7. Complete all bases with conjectured k < 50K to n=25K; [COLOR=red]completed 2018/02/09.[/COLOR]

8. Complete new bases with conjectured k < 100K to n=10K; [COLOR=red]completed 2018/01/17.[/COLOR]

These are agressive goals but with everyone's help I'm confident that we can accomplish many or all of them. If anyone would like to add a different goal or a different target for one of the above, please feel free to discuss.

In the future, we will plan to set such goals around the beginning of each year.

masser 2018-01-04 19:24

Alternate Goals Suggestion
 
I count 59 (announced) top 5000 primes in 2017 in this thread:

[URL]http://mersenneforum.org/showthread.php?t=9782[/URL]

100 top 5000 primes in 2018; [COLOR=red]completed 2018/05/12[/COLOR][COLOR=seagreen][COLOR=green].[/COLOR][/COLOR]
50 top 3000 primes in 2018; [COLOR=green][COLOR=red]completed 2018/05/01[/COLOR].[/COLOR]
25 top 1000 primes in 2018;[COLOR=green] 8 primes to go.[/COLOR]
10 top 100 primes in 2018;[COLOR=green] 10 primes to go.[/COLOR]
3 top 50 primes in 2018;[COLOR=green] 3 primes to go.[/COLOR]

Any of those potential goals sound appealing?

rogue 2018-01-04 20:10

[QUOTE=masser;476415]I count 59 (announced) top 5000 primes in 2017 in this thread:

[URL="http://mersenneforum.org/showthread.php?t=9782"]http://mersenneforum.org/showthread.php?t=9782[/URL]

100 top 5000 primes in 2018?
50 top 3000 primes in 2018?
25 top 1000 primes in 2018?
10 top 100 primes in 2018?
3 top 50 primes in 2018?

Any of those potential goals sound appealing?[/QUOTE]

One or more of those does sound appealing, but I don't have a feel for any goals related to that.

MisterBitcoin 2018-01-05 17:48

All goals, expect #1, can be reached by us.
As you may have seen I started working on R109 and will test other bases as well to n=400K.

Sieve files for #4 and #5 may be hard to generate. I recommend doing the files for #5 at your own.
I´ll take this year 2-3 bases from n=10K-25K, depending how long they need.

gd_barnes 2018-01-05 21:47

[QUOTE=MisterBitcoin;476535]All goals, expect #1, can be reached by us.
As you may have seen I started working on R109 and will test other bases as well to n=400K.

Sieve files for #4 and #5 may be hard to generate. I recommend doing the files for #5 at your own.
I´ll take this year 2-3 bases from n=10K-25K, depending how long they need.[/QUOTE]

On #2, you and Yoyo have already sieved most of the Sierp bases (b<=736). I think that BOINC could complete testing many or all of them on both sides if you can complete the sieving. I would only recommend that BOINC test them to n=400K or 500K. That should not take them too long for each base. I think if we can complete testing of the 1kers to n=400K or 500K then the proofs for #1 will follow. There will also likely be a couple of proofs from #3.

I can do most of the sieving for #4. There are something like 10 bases that need sieving and only about 20 bases that need testing. About 10 of them have already been sieved. These are surprisingly large efforts for an individual so BOINC could do all of the testing. N=25K-100K seems to be their favorite testing range anyway.

I can easily do all of the sieving for #5. There are about 6 bases plus there would be any new bases that are currently being tested to n=10K that need to be sieved. The tests would likely be too short for BOINC so you, me, and others could work on those.

I'm not saying that we can complete all of these but we can accomplish most of them with these targets in mind.

This is my effort to finally give the project some direction. :-)

gd_barnes 2018-01-26 00:12

#8 was completed on Jan. 17th.

Only S522 to n=25K is needed to complete #7. :smile:

gd_barnes 2018-02-09 22:48

And #7 has now been completed! :smile:

rogue 2018-02-24 14:47

#6 has been completed, but needs verification from Gary. I'll likely need to run ranges of remaining k to make sure that I didn't miss any primes.

gd_barnes 2018-02-24 18:15

[QUOTE=rogue;480766]#6 has been completed, but needs verification from Gary. I'll likely need to run ranges of remaining k to make sure that I didn't miss any primes.[/QUOTE]

Very well done on a huge effort, Mark! :smile:

LaurV 2018-02-25 03:09

No "R66 to 100k" in the goals list? :sad:

VBCurtis 2018-02-25 04:05

[QUOTE=LaurV;480801]No "R66 to 100k" in the goals list? :sad:[/QUOTE]

I'm on it!! Sieve is nearly 30G, LLR is, uhm, leisurely. Wanna do a million tests or so?

OK, how about 10,500 to 11,000? I'm almost down to 100k k's. :)

masser 2018-02-25 04:21

1 Attachment(s)
A million tests? LOL...

gd_barnes 2018-02-25 08:06

[QUOTE=LaurV;480801]No "R66 to 100k" in the goals list? :sad:[/QUOTE]

[QUOTE=VBCurtis;480804]I'm on it!! Sieve is nearly 30G, LLR is, uhm, leisurely. Wanna do a million tests or so?

OK, how about 10,500 to 11,000? I'm almost down to 100k k's. :)[/QUOTE]

Please not. Thank you.

If you can spend half as much CPU time proving some of the 1k bases or pushing some of the 2k-3k bases a little higher, that would be much more appreciated and help the project far more.

LaurV 2018-02-25 12:06

Ha, Garry, got you! It was a kind of joke, hehe, as R66 is still my baby... (and now, Curtis' baby too, but I promise you we never have any contacts each other in the past :razz:, this is a miracle...)

I am not going to work any R66 now. Maybe later, June, etc., now all my resources are busy, and I am also cutting on the electricity consumption, all the money go south-east... (the little miss LaurV started second year of the Uni in Australia,... she stays upside down now, Down Under... where women glow and men plunder... yaay... kookaburra... )

rogue 2018-02-25 14:53

As the goals are met, please update the first post with some indication that the goal has been met along with a date that it was met. Maybe the ones met could be posted with a different color text.

gd_barnes 2018-03-07 19:55

I have revised goals 2 and 3. After doing a more detailed analysis I believe that they were unrealistic. Goal 2 has changed from all 1k bases to n>=400K to only 1k bases <= 700 to n>=400K. Goal 3 has changed from <= 4k bases to n>=200K to <= 3k bases to n>=200K.

I have also added a count down meter for the goals not yet done. :smile:

gd_barnes 2018-03-07 20:14

[QUOTE=masser;476415]I count 59 (announced) top 5000 primes in 2017 in this thread:

[URL]http://mersenneforum.org/showthread.php?t=9782[/URL]

100 top 5000 primes in 2018?
50 top 3000 primes in 2018?
25 top 1000 primes in 2018?
10 top 100 primes in 2018?
3 top 50 primes in 2018?

Any of those potential goals sound appealing?[/QUOTE]

BOINC et al has already put a big dent in this goals list. In 2018 we've already had:

66 top 5000 primes
48 top 3000 primes
3 top 1000 primes
0 top 100 primes
0 top 50 primes

Info. gathered from the top-5000 site. This counts the entry point of the primes. Some have dropped into lower categories including 3 that have dropped off the top-5000 list.

Based on our other goals, the bottom two are probably unrealistic but the top two will easily be accomplished and the 3rd one will likely be accomplished as we do more testing on 1k bases for n>200K.

I will add a count down meter in Masser's post that I will keep updated.

pepi37 2018-03-08 10:14

[QUOTE=gd_barnes;481816]BOINC et al has already put a big dent in this goals list. In 2018 we've already had:

66 top 5000 primes
48 top 3000 primes
3 top 1000 primes
0 top 100 primes
0 top 50 primes

Info. gathered from the top-5000 site. This counts the entry point of the primes. Some have dropped into lower categories including 3 that have dropped off the top-5000 list.

Based on our other goals, the bottom two are probably unrealistic but the top two will easily be accomplished and the 3rd one will likely be accomplished as we do more testing on 1k bases for n>200K.

I will add a count down meter in Masser's post that I will keep updated.[/QUOTE]

If I found only one prime in my two 4*53 and 4*155 sequence it will be ( currently) around 31-th place :)
So even last one is not so impossible

gd_barnes 2018-05-12 19:55

The top two top-5000 goals shown in post 2 have now been accomplished. A nice run of primes by BOINC! :smile:

gd_barnes 2018-07-07 18:47

Goal #5 has been completed. :smile:

All remaining new bases will have > 1000 k's remaining at n=10K.

gd_barnes 2018-08-29 22:14

Goal #4 has been accomplished. :smile:

gd_barnes 2018-12-01 22:16

Goal #3 has been completed. :smile:

masser 2018-12-31 20:23

2019 project goals
 
[QUOTE=masser;476415]I count 59 (announced) top 5000 primes in 2017 in this thread:

[URL]http://mersenneforum.org/showthread.php?t=9782[/URL]

100 top 5000 primes in 2018
50 top 3000 primes in 2018
25 top 1000 primes in 2018
10 top 100 primes in 2018
3 top 50 primes in 2018

Any of those potential goals sound appealing?[/QUOTE]

It was cool to see all of the primes found by CRUS this year. I hope I am a little more lucky next year.

Progress on these goals was a little like Goldilocks and the three bears. TBH, I hadn't closely examined the project's history when I picked the optimistic targets above. More realistic and informed (but still challenging!) goals for 2019 might be:

100 top 3000 primes in 2019
20 top 1000 primes in 2019
5 top 200 primes in 2019

Do we like top-5000-list-based goals?

gd_barnes 2019-01-01 09:58

[QUOTE=masser;504521]It was cool to see all of the primes found by CRUS this year. I hope I am a little more lucky next year.

Progress on these goals was a little like Goldilocks and the three bears. TBH, I hadn't closely examined the project's history when I picked the optimistic targets above. More realistic and informed (but still challenging!) goals for 2019 might be:

100 top 3000 primes in 2019
20 top 1000 primes in 2019
5 top 200 primes in 2019

Do we like top-5000-list-based goals?[/QUOTE]


Sure! Those sound good. I'll come up with a new 2019 goals list comparable to the 2018 list over the next week.

gd_barnes 2019-01-03 09:47

We made a lot of progress in 2018 accomplishing the goals that were set at the beginning of the year. We accomplished 6 out of 8 goals set in the first post and 2 out of 5 goals set in the second post. :smile:

Look for a new thread with 2019 goals shortly.


All times are UTC. The time now is 08:06.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.