Gary,

Here's my humble opinion.
You shouldn't be worried about TPS large number of primes because in my point of view quality of primes is better than quantity. You should focuse on your 13 heavy-weights k's. What's is worthwhile is to pick up a fixed k and go up by your self and not doing us they are doing.
BTW, I am having lots of fun in here looking for some cool primes and I am already looking for a new quad machine to bring to the party. Thank you guys.

Best Regards,

Carlos

Thanks and note about RAM for new machines
 

[quote=em99010pepe;112039]Gary,

Here's my humble opinion.
You shouldn't be worried about TPS large number of primes because in my point of view quality of primes is better than quantity. You should focuse on your 13 heavy-weights k's. What's is worthwhile is to pick up a fixed k and go up by your self and not doing us they are doing.
BTW, I am having lots of fun in here looking for some cool primes and I am already looking for a new quad machine to bring to the party. Thank you guys.

Best Regards,

Carlos[/quote]

Carlos,

Thanks for the input. I definitely agree with you on quality vs. quantity. Fortunately in my case, I can accomplish two things at once. Taking all of my k up to top-5000 level while at the same time knocking off the 'lower-quality' n=333333 primes and ultimately not leaving gaps behind. Also if I start with a few k's on the side that are already tested up to n=250K to 400K or so, then I can feel OK about not leaving gaps behind on those while still finding top-5000 primes fairly quickly.

Bring on that new quad machine! :smile: That'd be nice to have for our effort here. One note on RAM if you do buy one...I need to go buy a GB or 2GB of RAM here in the next few days. My k's are so heavy-weight and numerous that the huge 'Legender symbol' file that is created and needed by sr2sieve is too large in order to allow it to run twice on my newer dual-core laptop with 1GB of RAM. I determined that this happens because it is running Windows Vista, which is a resource hog. It runs just fine twice on my slightly older dual-core main desktop that also has 1GB of RAM but is running Windows XP. Stupid operating system! :wink:

I bring this up because if you're getting a quad machine, be sure it has AT LEAST 2 GB or even 4 GB of RAM so that you can take full advantage of the 4 cores with all of your efforts here.


Gary

new data pages before holiday
 

hi all,
i sent new data pages to Kosmaj:
- new Woodall prime included (woodallprime.htm and in the summary)
- updated page with contributors
- updated page with programs (srsieve new versions)
- updated InfoSummary.htm

new:
- marked twins: k*2^n-1 and k*2^n+1 in k<3009 (not all yet) and in summary (data from Gary Barnes, thanks)

i also got a list with Riesel numbers from Wilfrid Keller and confirm the entries i found in the Top5000 as Riesel numbers: 18 k's with first prime between 200k and 250k. more to come (all k's with first prime > n=8192. about 320 k's).

so i wish everyone a fine holiday too.
i wonder how many primes to find in these 2 weeks. and some work for me to do!
greets Karsten

The above, latest version of stats files is now on the server.

For k=151515, following twin primes are missing: 43, 86, 114.
For k=1515, there is a twin at n= 105. Also there is a Cunningham chain of the second kind: 1515*2^105+1 (2p-1) :wink:

notes on twins
 

[quote=Cruelty;112285]For k=151515, following twin primes are missing: 43, 86, 114.
For k=1515, there is a twin at n= 105. Also there is a Cunningham chain of the second kind: 1515*2^105+1 (2p-1) :wink:[/quote]

Cruelty,

Karsten has only completed the notation of twins for the range of k = 301 to 585, k = 3011 to 10000, k = 1000065, and the largest k on the site. For almost all of them, I think he is using my twin list for k < 100K and n < 10K that is posted in the twin prime search forum on this site so I'm sure he would get to the one on for k=1515 pretty quickly after getting back from vacation. (I haven't checked Cunningham Chains.)

A couple of more notes about this:
1. The ones that you showed for k = 151515 are above my list so those are new twins that he wouldn't be aware of so those help us here.
2. If you want to additionally aid in the effort; for any k > 10K that you have previously found primes for, if you can search them for twins and post them here, then that would be helpful. I have all of my k searched and will post the twins for them when he gets back.


Karsten,

I did find some twins that were missing in the range of k that you did update for twins that are listed in my large twins list:

k = 573; n = 344
k = 7755; n = 1367
k = 9345; n = 445
k = 1000065; n = 26 (since you already listed one for 1000065, I checked them all for it and found this additional one)


One more thing...for k = 151023, it should show as tested up to n = 160K. Currently it doesn't show a testing limit.


Thanks,
Gary

Slight correction
 

I stated the following that was slightly incorrect:

[quote=gd_barnes;112343]
If you want to additionally aid in the effort; for any k > 10K that you have previously found primes for, if you can search them for twins and post them here, then that would be helpful.
[/quote]

This should have stated that anything for k > 100K would be helpful instead of k > 10K. My post in the twin prime search forum is for k < 100K and n < 10K.

Of course if you happen to be lucky enough to find one for n > 10K, regardless of the size of k, that would be REALLY helpful! They are very rare and hard to find unless you do a concentrated effort on a single n with multiple k. I'm currently doing an effort to take the list up to n = 15K and I've had to expand it to k < 1M because I was only getting 1-2 twins per 1000 n for k < 100K. The effort is up to about n = 13500 now. (Note that I'm not searching up to k < 1M for n < 10K. There are too many of those.)


Gary

I think I'll start today work on 100001<k<199999, 1<n<20000, however I will not focus on twins or SGs, afterwards I will run the resulting primes through Proth to find all twins and SGs :flex:

Clarification on range
 

[quote=Cruelty;112392]I think I'll start today work on 100001<k<199999, 1<n<20000, however I will not focus on twins or SGs, afterwards I will run the resulting primes through Proth to find all twins and SGs :flex:[/quote]

Excellent. I agree that when looking for twins or SG's across a large range of k, it's easier to search for all Riesel primes and then search those primes for Proth's or Riesel's for n+1.

As a clarification, will you be searching ALL k or all of the k on the summary page or just your own k in that range? I ask because eventually I hope to search for twins on ALL primes on our entire summary site or coordinate an effort to do so. :flex: Although that sounds like a monsterous task, twin searches on already found primes generally go very fast because most of them are sieved out before LLRing.

You may already be aware of this but if not, it might be helpful to know. When searching for twins, you can eliminate any k that is not divisible by 3.


Gary

I will test every k in the specified range till n=20000 :cool:
It will probably take me a week or even more, but what the hell... as I have already mentioned I am not particularly looking for twins or SGs, such a range of tested "k" might be usefull to choose some new candidates for team efforts based on computed weight :tu:

One more question...
 

[quote=Cruelty;112411]I will test every k in the specified range till n=20000 :cool:
It will probably take me a week or even more, but what the hell... as I have already mentioned I am not particularly looking for twins or SGs, such a range of tested "k" might be usefull to choose some new candidates for team efforts based on computed weight :tu:[/quote]


Wow, really cool! :cool: :grin: It's great to hear someone who thinks like me.

One more question...Do you plan to list ALL primes in this forum of any kind for the range that you mentioned like I did for k < 10K and n < 10K? If so, I'd suggest starting at k = 10K since that's where I left off at on my 'regular' prime search that Karsten has now posted. That way, there'd be no gap in k-values in the listing. Also, be sure and let Karsten know well ahead of time since it is a huge effort for him to get them all into the correct format for posting on the web pages. I really didn't give him enough notice on my effort.

But if you're only going to list twins and SG's in this forum, then doing so for k > 100K is definitely the best way to go to avoid duplicating efforts. My 'slow-computer' search for ONLY twins is up to around n = 14K for k < 1M.

Karsten or Kosmaj, if you're reading this; don't worry about my effort on twins. There will probably only be ~100 twins for the range of 10K < n < 15K and k < 1M that I am currently working on. And many of them will be for k's that aren't currently on the summary site. I'm guessing you'd only want to note the twins on our site if the k has already had a regular prime search done on it so the effort from your end should be relatively small. I just want to keep a comprensive list of twins in one place for future use; that place being the twin prime search forum on this site.


Gary