PrimeGrid work on 90646*2^n + 1 and 101746*2^n + 1

JeppeSN · 2020-06-30, 21:06

In the context of the PrimeGrid forum, the idea to find additional primes in the series 45323*2^n + 1 and
50873*2^n + 1, along with others, came up. They both produce "trivial" primes for n=1.

These were searched "manually" at first, and later searched coordinated in a private BOINC server, while testing new software (LLR2, based on LLR).

We have now become aware that Conjectures 'R Us are searching the same series, written as 90646*2^n + 1 and 101746*2^n + 1 (so multipliers doubled, the trivial primes now occur at "new" n=0).

I just wanted to make you guys aware of this work duplication. See my recent post https://www.primegrid.com/forum_thre...ap=true#141293 for more details.

/JeppeSN

gd_barnes · 2020-06-30, 22:18

Quote:

Originally Posted by JeppeSN

In the context of the PrimeGrid forum, the idea to find additional primes in the series 45323*2^n + 1 and
50873*2^n + 1, along with others, came up. They both produce "trivial" primes for n=1.

These were searched "manually" at first, and later searched coordinated in a private BOINC server, while testing new software (LLR2, based on LLR).

We have now become aware that Conjectures 'R Us are searching the same series, written as 90646*2^n + 1 and 101746*2^n + 1 (so multipliers doubled, the trivial primes now occur at "new" n=0).

I just wanted to make you guys aware of this work duplication. See my recent post https://www.primegrid.com/forum_thre...ap=true#141293 for more details.

/JeppeSN

Hmmm...Can you provide a link to the BOINC server? What is the test limit?

We are also searching Riesel k=351134 and k=478214 in the same manner.

Our test limit is currently at n=6.6M on both the Riesel and Sierp even k's and continuing on our public PRPnet server.

JeppeSN · 2020-06-30, 23:12

Quote:

Originally Posted by gd_barnes

Hmmm...Can you provide a link to the BOINC server? What is the test limit?

We are also searching Riesel k=351134 and k=478214 in the same manner.

Our test limit is currently at n=6.6M on both the Riesel and Sierp even k's and continuing on our public PRPnet server.

Not sure about the test limit. The BOINC server is at http://boincvm.proxyma.ru:30080/test...ver_status.php, the subproject has the name "LLR2 testing", but I do not think the search limit it visible. There was one reportable prime found, namely 68633*2^2715609 + 1.

It must be the case that your PRPNet server is ahead of the private BOINC server on these two k. I think we are going to regard the work on the private BOINC server as a double check of your work (for these two multipliers).

/JeppeSN

stream · 2020-07-01, 07:00

Quote:

Originally Posted by gd_barnes

Hmmm...Can you provide a link to the BOINC server? What is the test limit?

We're searching k=23971, 45323, 50777, 50873, 76877 for Proth primes because currently known primes for these k's do not match Proth criteria 2^n > k. We do not search Riesel forms.

Server link is boincvm.proxyma.ru:30080/test4vm/ but this search has no dedicated subproject. We loaded these candidates into existing "LLR2 testing" subproject to test our software and server limits. These 5 k's are currently tested up to 4M and search from 4M to 7M is planned to be started in middle of July. An announce will be done on project thread https://www.primegrid.com/forum_thread.php?id=9107

Is your work double-checked? If not, we can just double-check your Proth work for these two k's up to 7M. If yes, we can step out and do not test these two k's anymore. Note that we're using LLR2 software which can generate proof of correctness and do not requires full double-checking.

gd_barnes · 2020-07-01, 18:39

No our work is not double-checked. That would be great if you double-check it.

gd_barnes · 2020-07-03, 02:47

So that you know where we are at on these k's, here is a link to our PRPnet server status page:

http://www.noprimeleftbehind.net:1400/all.html

JeppeSN · 2020-08-05, 17:12

That private BOINC server now uses the residues from CRUS for these two k, and double checks your work. And it has a status page now: http://boincvm.proxyma.ru:30080/test...r2_status.html

Thanks to Stream.

/JeppeSN

chalsall · 2020-08-05, 17:45

Quote:

Originally Posted by JeppeSN

Thanks to Stream.

Hey... Do you think Stream would be willing to provide some quick BOINC wisdom?

If so, please PM me and I'll provide my email address. Shouldn't take much of his time.

stream · 2020-10-17, 14:40

We finished Proth testing of k=23971, 45323, 50777, 50873, 76877 (*2^n+1) up to n=10M, using trusted computation (LLR2 proof of correctness A.K.A. fast double-check). Unfortunately, no primes were found.

kar_bon · 2020-10-19, 10:45

I've included those 5 k-values in the Wiki with current status and a project page (need more history/text).

gd_barnes · 2020-10-20, 02:53

Info for CRUS folks: Sierp base 2 k=90646 and 101746 have been completed to n=10M and are being released. Stream will send me the residues in a few days.

When other k's in our Riesel/Sierp base 2/4 even-odd k-n effort that are being run on our PRPnet server port 1400 are complete to n=10M CRUS will pick them back up.

2020-10-17, 14:40	#9
stream May 2015 4₁₆ Posts	We finished Proth testing of k=23971, 45323, 50777, 50873, 76877 (2^n+1) up to n=10M, using trusted computation (LLR2 proof of correctness A.K.A. fast double-check). Unfortunately, no primes were found. Last fiddled with by stream on 2020-10-17 at 14:41*

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
The Primegrid monopoly	The Carnivore	Riesel Prime Search	9	2011-04-24 20:55
ROLP at PrimeGrid	mdettweiler	Raiders of the Lost Primes	2	2010-09-17 05:57
PrimeGrid site down :(	mazadillon	Prime Sierpinski Project	2	2008-12-24 00:19
Looks like PrimeGrid's noticed us....	mdettweiler	Conjectures 'R Us	7	2008-03-06 17:33
Anyone knows about primegrid	koders333	Factoring	8	2006-03-14 03:23

2020-06-30, 21:06	#1
JeppeSN "Jeppe" Jan 2016 Denmark 2²×41 Posts	*PrimeGrid work on 906462^n + 1 and 1017462^n + 1* In the context of the PrimeGrid forum, the idea to find additional primes in the series 453232^n + 1 and 508732^n + 1, along with others, came up. They both produce "trivial" primes for n=1. These were searched "manually" at first, and later searched coordinated in a private BOINC server, while testing new software (LLR2, based on LLR). We have now become aware that Conjectures 'R Us are searching the same series, written as 906462^n + 1 and 1017462^n + 1 (so multipliers doubled, the trivial primes now occur at "new" n=0). I just wanted to make you guys aware of this work duplication. See my recent post https://www.primegrid.com/forum_thre...ap=true#141293 for more details. /JeppeSN

2020-07-01, 18:39	#5
gd_barnes May 2007 Kansas; USA 282D₁₆ Posts	No our work is not double-checked. That would be great if you double-check it.

2020-07-03, 02:47	#6
gd_barnes May 2007 Kansas; USA 5·11²·17 Posts	So that you know where we are at on these k's, here is a link to our PRPnet server status page: http://www.noprimeleftbehind.net:1400/all.html

2020-08-05, 17:12	#7
JeppeSN "Jeppe" Jan 2016 Denmark 2²×41 Posts	That private BOINC server now uses the residues from CRUS for these two k, and double checks your work. And it has a status page now: http://boincvm.proxyma.ru:30080/test...r2_status.html Thanks to Stream. /JeppeSN

2020-10-19, 10:45	#10
kar_bon Mar 2006 Germany B32₁₆ Posts	I've included those 5 k-values in the Wiki with current status and a project page (need more history/text).

2020-10-20, 02:53	#11
gd_barnes May 2007 Kansas; USA 5·11²·17 Posts	Info for CRUS folks: Sierp base 2 k=90646 and 101746 have been completed to n=10M and are being released. Stream will send me the residues in a few days. When other k's in our Riesel/Sierp base 2/4 even-odd k-n effort that are being run on our PRPnet server port 1400 are complete to n=10M CRUS will pick them back up.