The simultaneous record page was a new one for me, thanks for letting us know.
- and congratulations with your new 10!
vidarn

Kudos to you!
And you deserve it - you wrote the program!
Keep up the good work:  Chasing primes and playing the Riecoin Game!
regards,
vidarn

This may not be strictly riecoin and 6-tuplet stuff, but I still think this forum is the best place to post it:

rieMiner 0.9 (by pttn) set a world record yesterday, finding a 527-digit 7-tuplet in Benchmark mode.
http://anthony.d.forbes.googlepages.com/ktuplets.htm
The record attempt ran for about 1130 hours on a Ryzen 1700.

regards,

vidarn/Art3mis

I have played a lot of games with Riecoins and 6-tuplets over the last five years, and when a new miner comes out, it certainly deserves to be tested. My initial experiences with rieMiner 0.9:

The binaries run without problems on both Intels and AMDs, using both Windows 7, 8 and 10 and Ubuntu 16.04. The source code compiles on my Ubuntu 16.04 -  without a hitch! Copy files, type make, and watch the build! Yes, I already had the prerequisite libraries and stuff.  And my compiled code runs just as well as pttn's binaries. I haven't tried a Windows build - yet.

On my Ryzen 1700, benchmarks show about 20% performance increase compared to fastrie. In actual mining, rieMiner 0.9 seems to do even better than that, although it is difficult to find equal circumstances for comparison.

Pttn and Rockhawk have done a great and thorough job rewriting the miner. For me, penetrating the new code will be a challenge, though. Personally, I prefer the procedural programming style of Jh00 and dga and get slightly nauseous whenever I meet double colons, namespaces, classes and suchlike.

But - disregard any negative thoughts from a grumpy oldtimer. Obviously, if someone completely rewrites, improves and restructures an existing system, he should be allowed to use whichever paradigm he prefers or finds necessary. Well done!

I usually don’t mine riecoins any more. It is probably cheaper to buy them than to mine them these days.  This Christmas however, as a small token of appreciation, I have fired up all the systems I have available (connecting to ublock.it) and have directed all the output to pttn's payout address:

RPttnMeDWkzjqqVp62SdG2ExtCor9w54EB

And folks, have a closer look at that address! I didn't know you could do that - Cool!

Merry Xmas to you all

vidarn

For the record, I tested the 745 digits candidate 1401767259 . . . 99497551 and it's 5 siblings using Primo, and they all came out with flying colors.

Vidar

Ahem...

745 is definitely the largest 6-tuplet ever found by the Riecoin system and may qualify for the term  record,
but the World Record is still the 1037 digits 6-tuplet found by Norman Luhn in March 2016.
(and will probably be so for a long time)

http://anthony.d.forbes.googlepages.com/ktuplets.htm

But it is really good to see ordinary difficulties in the 1700s again, brings back memories from May 2014!

regards,
Vidar

I got in touch with Norman Luhn, and he gave me this brief:

The titanic-6-tuplet-project was a "dream" of me in year 2000.
I have started the project on Aug. 2011. After 4 years I found it after 153 quintuplets. The hardware is normal, 12-18 Phenom II cores at 2.8 GHz. I wrote a presieve for APSIEVE. It was a hard job :-) to sieve in 10,000 4-billion-blocks before I found it. I looked for k's in an interval 0 up to 10^14.

Since he just signed up for this forum, you can probably discuss other details with him directly.

Hi all,
The 6-tuplet world record has been broken again, thoroughly:

The previous record was 751 digits, Norman Luhn's new record has 1037 digits.

The first titanic (more than 1000 digits) 6-tuplet record is
28993093368077 * 2400# + 19417
http://anthony.d.forbes.googlepages.com/ktuplets.htm

Norman is a veteran prime record breaker and held a lot of the previous 6-tuplet records.

The number above corresponds to a Riecoin difficulty of 3444.
Wow!

Xpoolx and port 5000 works quite nicely for me. Also, within the last 30 minutes, I solved a couple of blocks on gpupool.com, which also works ok.
The problem may be on your side of things?

The only dedication necessary is pressing Enter – and wait - and make sure you have enough spare machines to do other important things in the meantime.

I am still using basically the same setup as with the previous records – a stripped version of dga’s B14 Riecoin miner. I estimate this system to be between 4 and 8 % faster than the B14 since I am only looking for 6-tuplets and abandon further primality tests as soon as I meet the first non-prime.

I was never able to tame the B15s and the B16s into the Microsoft Visual Studio harness, discouraged by several hundred compile time challenges. (The B14 only gave me about a dozen)

My present tuplet searcher is probably just on par and maybe even a bit slower than the B16 miners (which I believe are the current best ones?)

I just had a quick look at the block times for the Hall of Fame:

superblock       datetime                   prev block datetime          block time           diff           digits
289872          03.07.2015 23:24          03.07.2015 21:32          01:51:32          2256.00          679
281808          19.06.2015 17:09          19.06.2015 16:26          00:42:27          2232.00          672
293904          10.07.2015 23:16          10.07.2015 23:02          00:13:42          2211.00          666
285840          26.06.2015 21:17          26.06.2015 21:16          00:00:50          2196.00          661
306000          01.08.2015 16:02          01.08.2015 15:30          00:31:11          2195.00          661
297936          18.07.2015 10:45          18.07.2015 10:17          00:27:56          2183.00          657
310032          08.08.2015 19:25          08.08.2015 18:10          01:14:27          2176.00          655
164880          24.11.2014 20:28          24.11.2014 16:13          04:14:39          2173.00          654
301968          25.07.2015 13:33          25.07.2015 13:12          00:20:45          2169.00          653
201168          27.01.2015 04:07          27.01.2015 02:40          01:26:12          2154.00          649

They average just above one hour (notice the 26 June entry, less than one minute!)

Using gatra’s formula from almost a year back: to estimate time for a block of a certain difficulty based on numbers from another diff:

T1/T2 = (Diff1/Diff2)^9

Based on the superblock entries above: Working at diff 2500 rather than 2200: ~3 times longer time or just over 3 hours on average.
Based on the present regular blocks: Working at diff 2500 rather than 1500: ~100 times longer time, just over 100*2,5 minutes  or around 4 hours.

Looking at the spread in the Hall of Fame, you would expect some of the possible 2500 superblocks to be solved in half an hour, but you could also expect some well over 10 hours. It could be done, or…? Also, back in June 2014 with a difficulty of 1700+ and well over 20k workers at ypool, an 800 digit (diff 2656) world record would be well within reach.

I don’t really know why I am encouraging this strange mining brotherhood *) to break my records – it is probably because I was a member myself from the very beginning – and occasionally still am.

Regards,

Vidar

*) and sisterhood – apologies to Cinnamon!

Thanks for the kind words, fairglu!

7 out of 8 cores of an AMD FX8350 running at normal speed (4 GHz)
The process took around 1920 hours, from May 14th to August 2nd

(When I did 700 in January, I had a stroke of luck and found a solution in under three days!)

Now, if you had 1920 such systems and somehow forced the superblock difficulty to 2500, how many hours would it take to find one 752 digit 6-tuplet?
Just a thought...

regards,
Vidar

Actually, I have raised the bar even further, 751 digits on the 8th of August:

http://anthony.d.forbes.googlepages.com/ktuplets.htm

To beat this with the present fork of Riecoin, you need a superblock with difficulty around 2495, corresponding to a general difficulty of 1706 or above. (Apparently, the superblock difficulty is about 1.462 times the normal difficulty.)

It has been a year or so since the difficulty was in the 1700s but both difficulty and exchange rate made some interesting jumps in June and July this year, so who knows…

But I have to admit, considering that the record 6-tuplet a year ago was 593 digits, the Riecoin Hall of Fame is really impressive!

Thanks!

Hi j0nn9,
In crt-rev5.zip I can only find the original source code (minus the POW) from October last year. Did you mix up some batches of code? (or am I doing something wrong?)

Wow! Imagine being on a computing merit list next to Donald E. Knuth!
Do you guys realize what an honor that is??

Congratulations!
Saw the confirmation at Tony Forbes’ today (http://anthony.d.forbes.googlepages.com/ktuplets.htm) – I am beaten by 33 digits! Great achievement, gatra and the crowd (and also libero2014, who happened to do the actual work)

Regards, Vidar

Thank you!

I am quite happy with my flavor of b14.  I do some fiddling with prime numbers occasionally  and spent a lot of my spare time this summer trying to figure out the sieve. (Part 2 of Fast Prime Cluster Search could have saved some time, but it was an interesting exercise tracking the ingenious code!)

I left out all the xpt and Riecoin block handling, the starting target was simply 2^difficulty,  with a smaller random number added.  With no Riecoin restrictions, I could use larger primorials. My only other change  to the sieve was  when I check the five remaining tuplet members for primality, I break the loop at the first non-prime. I know I won't get a full  6-tuplet and searching for a 4 or 5-chain with a "hole" in it is of no use. The mpz_powms in the Fermat tests are quite time consuming!

BTW, while I was thinking about where to go next *) the program was idling away here the other day and (almost accidentally!) found a 600-digit 6-tuplet. I'll stop there and leave the rest to the Riecoin Superblock.

Regards, Vidar


*)  Fortunate numbers at present,  definitely not Riecoin stuff, probably both impossible and useless, but fun!

Hi,
A few days ago I submitted to Anthony Forbes a 597-digit 6-tuplet and had it verified as a new world record:

http://anthony.d.forbes.googlepages.com/ktuplets.htm

This number is only a few digits above the previous record, and it should be quite possible for the superblock project to beat it!

The program used was based on the b14 version of dga's xptminer (which was the latest version I was able to compile using Windows and Visual Studio) and ran on 11 cores for about a month before it suddenly surprised me with a find. I'll give some further details later.

The 6-tuplet is

1268360451482944051031693571608646702803337857417059923179003396370803464447945 9213486050679397935104622682138475913438101505200313946821621485384179510473697 9433691754086534863323734928743959323221865529501237655633762351765417920461874 1258860678645822691792395092270199088414732238122470003707838866453023159582123 7351808046531772170786266328462890225388980932631840153394721839158254583153068 7931209705774579874734193952405536748117312631969874545797052973507394025122061 8770487542157843118646752735712130590319004160226700306792249093957168851751545 30232372583285744387264936607664813582967597
+d, d = 0, 4, 6, 10, 12, 16

or

4921035090333221245281056600469203091910280140820896181940314619657756374547612 4680711202733554189605614783506825475691404888752144184576829642013820327715134 5176983584805931408105542588834038851652656017513154942170087395616822290168504 6046380856370538924059371410139041704192384855299316496758299832292941057155474 6087497007967825846220015325654831770578435398632279854720 * 547# + 8061997 +d, d = 0, 4, 6, 10, 12, 16

Regards,
Vidar

Unlike Riecoin, Primecoin makes it possible to submit POW well above the current difficulty. The present Primecoin record is a 2nd kind Cunningham chain of length 14 from May this year (when the difficulty was 10.96)

http://primecoin.io/index.php
http://primerecords.dk/Cunningham_Chain_records.htm

I believe you guys are making your estimates based on an old world record.
The 1857 difficulty corresponds to a 559 digits 6-tuplet found in 2009. The present record, according to Tony Forbes, is 593 digits or a difficulty of about 1970. The record 6-tuplet is

219946485329 * 1399# / 2 + d, d = −8, −4, −2, 2, 4, 8, found by Serge Batalov in December 2013.

Tony Forbes' k-tuplet page may be found here:

http://anthony.d.forbes.googlepages.com/ktuplets.htm

But as (1973/1441)^9 = 16.9 using gatra's estimates, we should be able to beat the record with an extra digit within an hour or so. The Primecoin lot have broken quite a few world records, it would be nice if we could get some too.