mersenneforum.org Improved params files for CADO
 Register FAQ Search Today's Posts Mark Forums Read

 2022-08-11, 15:51 #100 VBCurtis     "Curtis" Feb 2005 Riverside, CA 15AF16 Posts When I first discovered that the stock params were not very fast, I sent new files to the CADO folks for C100 to C120. Some of them were accepted, others were deemed not a useful improvement. One of the developers explained that the team uses an automated params testing suite to determine the stock params, and I got the sense they were not particularly interested in my efforts to find faster ones. Quite a few years have passed since then, and I've learned a lot about parameter choices. Perhaps I should send the team my most-improved params files once again, as I've found more and more speed over the years.
 2022-08-15, 12:24 #101 bur     Aug 2020 79*6581e-4;3*2539e-3 601 Posts Yes, I thought that maybe they focus more on the development of the software and bleeding edge parameters than those smaller ones. Still, it would be very interesting for publication (not peer-reviewed, just to make it public). Will you post the updated c155 params shortly? Not to sound impatient, I just want to make sure I don't start factorization and then 3 hours later it's posted. :) Last fiddled with by bur on 2022-08-15 at 12:25
 2022-08-15, 14:18 #102 VBCurtis     "Curtis" Feb 2005 Riverside, CA 7·13·61 Posts I'm working this week on C140 and C145; I seem to have lost the trendline here, having trouble finding settings fast enough to match my expectations based on C120 and C130 timings. I won't be publishing C150 and C155 for a while still.
 2022-08-16, 05:36 #103 bur     Aug 2020 79*6581e-4;3*2539e-3 601 Posts Ok, good to know.
 2022-09-15, 15:41 #104 storm5510 Random Account     Aug 2009 Not U. + S.A. 24·149 Posts I still use all the stock parameters from the original installation which was recently. A document refers to this version as 3.0.0. I have all the newer parameter files posted here on a USB drive. Should I replace the original ones or leave them alone?
 2022-09-15, 17:50 #105 VBCurtis     "Curtis" Feb 2005 Riverside, CA 7·13·61 Posts You should get the 2022 parameters from the thread so marked, and replace your stock files. I mean, unless you prefer jobs taking 10-40% longer than they need to- it's not like you must replace the parameters files to run the program. Some of the factory CADO params are fine, like C110; others have weird choices that cost a ton of speed. In most cases, the poly select params I use make jobs about 2 digits easier than the factory settings; that's the bigger boost than sieve params.
 2022-09-17, 14:19 #106 storm5510 Random Account     Aug 2009 Not U. + S.A. 24·149 Posts I just happened to see the warning below as it appeared on the screen. This was after the parameter file updates. Code: Debug:HTTP server: 127.0.0.1 Translated path cgi-bin/upload.py to /home/levi/cado-nfs/scripts/cadofactor/upload.py Info:HTTP server: 127.0.0.1 Sending workunit c130_sieving_5040000-5060000 to client localhost+2 Debug:HTTP server: 127.0.0.1 "GET /cgi-bin/getwu?clientid=localhost+2 HTTP/1.1" 200 - Warning:Command: Process with PID 862234 finished with return code 2 Info:Filtering - Singleton removal: After purge, 7583130 relations with 7388066 primes remain with excess 195064 Info:Filtering - Singleton removal: Not enough relations Info:Filtering - Singleton removal: Requesting 1735649 additional relations Info:Filtering - Duplicate Removal, removal pass: Got request for 29512333 (1735649 additional) output relations, estimate 42565082 (2503293 additional) needed in input Info:Lattice Sieving: New goal for number of relations is 42565082, currently have 40061789. Need to sieve more Debug:Filtering - Singleton removal: Exit PurgeTask.run(purge) Info:Complete Factorization / Discrete logarithm: Filtering - Singleton removal Info:Filtering - Singleton removal: Total cpu/real time for purge: 77.42/80.55 Info:Lattice Sieving: Starting Much of the time, I would have never seen this as I wouldn't be paying any attention or the monitor would be turned off. I may have gotten it before and not known about it. CADO properly finished the process.
2022-11-03, 12:46   #107

Apr 2022

22 Posts

Quote:
 Originally Posted by orever Is this another case of a secret number or can you say the number? Maybe with ECM you can factor faster, or is it an RSA key?
Hi Sorry for the late super reply. Yes it is a RSA key but we are still testing our software offline so I can post our testing PK. The only problem is if we make a release with a 640bit key...

Our testing public key is:
Code:
N:3258005875255212804435711892724048997709841623079090536527862146732351426484056684475910406742090087618411947018308228277114851649885017479304690479085490775846701637841220085809920785148488781
E: 65537
I haven't had any lock with getting Faas to work again. And it is sadly not in the factordb!

Last fiddled with by VBCurtis on 2022-11-03 at 20:06 Reason: added code tags

 Similar Threads Thread Thread Starter Forum Replies Last Post Xyzzy Forum Feedback 3 2018-12-30 19:37 jasonp Operation Kibibit 5 2014-09-07 11:02 skan Information & Answers 1 2013-10-22 07:00 jasong GMP-ECM 11 2007-05-30 03:08 eepiccolo Lone Mersenne Hunters 3 2003-04-12 02:04

All times are UTC. The time now is 20:57.

Fri Dec 2 20:57:18 UTC 2022 up 106 days, 18:25, 0 users, load averages: 0.85, 1.03, 1.09