Register FAQ Search Today's Posts Mark Forums Read

 2017-08-21, 10:47 #1 Aillas     Oct 2002 France 3×5×11 Posts Any updates please ? Hi, XYYXF website has not been updated since January. Is there an update soon planed? Web site deserted ? Thanks
 2018-01-04, 12:27 #2 Aillas     Oct 2002 France A516 Posts Hi, I know a lot of you are busy on lot of projects, but I would like to know if the XYYX web site is abandoned or not? The last update is 1 year old. Thanks and happy new year.
 2018-01-04, 14:04 #3 swellman     Jun 2012 3,511 Posts I do not know the status of the xyyxf website (Andrey the admin has been silent since last spring), though I am still actively working the project. Mostly providing additional ECM on numbers after Yoyo@Home does the heavy lifting and then feeding them to NFS@Home for factoring. I do keep track of the project in my records. There is much work to be done if you are interested. PM me.
 2018-01-05, 09:41 #4 Aillas     Oct 2002 France 16510 Posts Thanks for the information. I'm already working on XYYX project using BOINC (Yoyo & NFS @home) projects. I don't have lot of cores and I'm switching to other projects as well from time to time. Just adding my drop in the ocean.
 2018-03-01, 21:24 #5 pxp     Sep 2010 Weston, Ontario 21410 Posts Just a note to point out that my Leyland prime indexing page is an alternative to the XYYXF website if your only interest is an up-to-date listing of known Leyland primes (sorted by size).
 2019-01-26, 11:46 #6 swellman     Jun 2012 1101101101112 Posts As of this morning 26 Jan 2019, ECM courtesy of Yoyo@Home is still grinding away. Yoyo has been running ECM @B1=110e6 for 18000 curves. Work at this level has been going since autumn of 2015. 344 composites remain to be factored in the xyyxf project. 99 composites have survived full t50. 232 composites have survived full t55. 13 composites have survived t60+. Yoyo is estimated to complete ECM of all composites to t55 by Sep 2019. Then ECM @B1=260M for 42,000 curves can begin. I estimate this t60 work will take 5-6 years for Yoyo to complete.
 2019-10-08, 08:28 #7 NorbSchneider     "Norbert" Jul 2014 Budapest 24·7 Posts How many composites remain in the xyyxf project?
2020-01-23, 12:52   #8
swellman

Jun 2012

351110 Posts

Quote:
 Originally Posted by NorbSchneider How many composites remain in the xyyxf project?
331 by my count. 4 are best solved using GNFS, the rest by SNFS.

Yoyo@Home is working through the last of the t55 ECM work, mostly some composite cofactors remaining after a prime factor was found. Should be done with that effort within 4 months.

Going forward, all remaining composites will require a minimum of t60 for ECM, and most will require partial or full t65. Since Yoyo limits the size of ECM jobs to < 20,000 curves for t60 and < 12,000 curves for t65, it will take repeated queuing to achieve the required levels of ECM before NFS methods can be employed. With that in mind, I’m planning on focusing ECM efforts on one small group of composites at a time until all ECM is complete and then passing them on to NFS@Home (or interested individuals with the necessary hardware). Repeat.

The above strategy will take many years to complete.

Also for the past 4+ years I have been running ECM (using my individual machines) on x=151 of the xyyxf form. 58 composites remain, and all but 3 have been tested to the t50 level. Anyone interested in these 58?

 2020-01-23, 13:15 #9 swellman     Jun 2012 66678 Posts One last thing - for years Andrey occasionally did a bit of high level ECM on the C3xx composites from this project. Only 11 of these C3xx remain. From Yoyo’s logs, as well as some private correspondence with individual contributors, I found 6 of these big composites have survived t65+, with the other 5 only having 9000 curves @t65. (If these 5 saw more ECM I can’t find any record of it.) My plan is to complete the t65 work on these last 5 big composites. This effort should be complete in a year.
2020-04-04, 16:35   #10
sparrow

Mar 2020

22 Posts

Quote:
 Originally Posted by swellman 331 by my count. 4 are best solved using GNFS, the rest by SNFS. Yoyo@Home is working through the last of the t55 ECM work, mostly some composite cofactors remaining after a prime factor was found. Should be done with that effort within 4 months. Going forward, all remaining composites will require a minimum of t60 for ECM, and most will require partial or full t65. Since Yoyo limits the size of ECM jobs to < 20,000 curves for t60 and < 12,000 curves for t65, it will take repeated queuing to achieve the required levels of ECM before NFS methods can be employed. With that in mind, I’m planning on focusing ECM efforts on one small group of composites at a time until all ECM is complete and then passing them on to NFS@Home (or interested individuals with the necessary hardware). Repeat. The above strategy will take many years to complete. Also for the past 4+ years I have been running ECM (using my individual machines) on x=151 of the xyyxf form. 58 composites remain, and all but 3 have been tested to the t50 level. Anyone interested in these 58? Comments?
Code:
yafu-x64.exe "factor(127^95+95^127)" -v -threads 8

fac: work done at B1=11000000: 0 curves, max work = 4480 curves
fac: 4480 more curves at B1=11000000 needed to get to t60.31
ecm: 4480/4480 curves on C196, B1=11M, B2=gmp-ecm default, ETA: 34 sec
fac: setting target pretesting digits to 60.31
fac: t15: 3840.07
fac: t20: 3269.11
fac: t25: 908.48
fac: t30: 224.55
fac: t35: 45.11
fac: t40: 7.69
fac: t45: 1.13
fac: t50: 0.15
fac: t55: 0.02
fac: sum of completed work is t45.73
fac: work done at B1=43000000: 0 curves, max work = 7553 curves
fac: 7553 more curves at B1=43000000 needed to get to t60.31
ecm: 567/7553 curves on C196, B1=43M, B2=gmp-ecm default, ETA: 245.83 hrs
I failed to factorize this number.

C196_127_95 How long will it take for this number to be factorized?
Attached Thumbnails

Last fiddled with by sparrow on 2020-04-04 at 16:37

2020-04-04, 21:42   #11
VBCurtis

"Curtis"
Feb 2005
Riverside, CA

17×311 Posts

Quote:
 Originally Posted by sparrow C196_127_95 How long will it take for this number to be factorized?
The ECM curves you're running aren't big enough to have a significant chance to find a factor. Try using B1=260M or bigger for a few days.

If GNFS is used for this number, figure 12 to 20 core-years (depending on one's "core") to do the factorization. How many cores do you have?
I didn't bother to check if SNFS if faster.

 Similar Threads Thread Thread Starter Forum Replies Last Post Baumann Eduard Information & Answers 5 2012-10-25 18:44 timespy Information & Answers 3 2009-04-28 20:26 Nichtwahr NFSNET Discussion 16 2007-04-27 14:45 ltd Prime Sierpinski Project 4 2007-02-09 02:21 SandStar NFSNET Discussion 7 2004-07-03 22:59

All times are UTC. The time now is 15:30.

Thu May 26 15:30:35 UTC 2022 up 42 days, 13:31, 2 users, load averages: 2.02, 1.35, 1.31