mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Factoring (https://www.mersenneforum.org/forumdisplay.php?f=19)
-   -   Python script for search for factors of M1277 using random k-intervals (https://www.mersenneforum.org/showthread.php?t=25941)

storm5510 2020-10-22 15:15

[QUOTE=VBCurtis;560649]Wait for what?

We haven't done nearly enough ECM to justify the time SNFS would take. You can wait for others to do the ECM, or you can contribute if you wish. I've done both, myself- I ran a couple CPU-years of ECM, and have waited since. I have a large-memory machine available now, so perhaps I'll restart a little large-bound ECM and contribute more than just posts.[/QUOTE]

[U]There is no need to be abrasive[/U].

I am running groups of these with [I]Prime95[/I] on the replacement machine using four workers. I have it set to [U]not[/U] report these to PrimeNet. Reason: I am not running these in a conventional way. I took the last B1 on the [I]mersenne.org[/I] table and doubled it, making 16e8. I will run it this way for a while, then increment it to 24e8, then 32e8, and so on, adding 8e8 as I go.

As for [I]GMP-ECM[/I], I do not know if I can keep it 'fed" with just four workers. If not, then I can run another instance of [I]Prime95[/I] on my primary machine, taking the total number of workers to eight. I would allow [I]GMP-ECM[/I] to pick its own B2 by not specifying it.

Lastly, it is a foregone conclusion some of you will say that I am wasting my time. The response is: It is my time to waste, not yours, so do not bug me about it...

ryanp 2020-10-22 15:59

[QUOTE=VBCurtis;560610]Ryan Propper doesn't always report his ECM work, so I would not be surprised to learn a full T70 or more has been completed. Similarly, I would be quite surprised if a factor below 69 digits turns up for this number.[/QUOTE]

I've unfortunately lost a complete count of curves, but I believe I did a full T65 as well as T70 on this number.

VBCurtis 2020-10-22 18:17

Thanks for confirming, Ryan!

T70 is often represented by curves at B1=2.9e9 or so. If that level is complete, then T75 curves are what is useful now- that's B1=7e9 or bigger. That's one curve per core per day range.... yeesh.

I'll give a curve at B1=1e10 a try, and see what ecm -v reports for B2, memory use, and # of curves to a T70 or T75.

ryanp 2020-10-22 19:04

[QUOTE=storm5510;560719]Lastly, it is a foregone conclusion some of you will say that I am wasting my time. The response is: It is my time to waste, not yours, so do not bug me about it...[/QUOTE]

More ECM on this number is not a waste, but it's becoming increasingly difficult. :smile: And it's recommended that you or anyone else do curves at B1=76e8 or higher, maybe 1e10 or 2e10 now.

bsquared 2020-10-22 20:26

[QUOTE=VBCurtis;560744]Thanks for confirming, Ryan!

T70 is often represented by curves at B1=2.9e9 or so. If that level is complete, then T75 curves are what is useful now- that's B1=7e9 or bigger. That's one curve per core per day range.... yeesh.

I'll give a curve at B1=1e10 a try, and see what ecm -v reports for B2, memory use, and # of curves to a T70 or T75.[/QUOTE]

I will run some tests using AVX-ECM. Scaled tests, at B1=7e6, show that first-stage curve throughput is about 2.4x larger than GMP-ECM.

[CODE]echo "2^1277-1" | ../../ecm-704-linux/ecm -v 7000000
GMP-ECM 7.0.4 [configured with GMP 6.2.0, --enable-asm-redc] [ECM]
Tuned for x86_64/k8/params.h
Input number is 2^1277-1 (385 digits)
Using special division for factor of 2^1277-1
Using B1=7000000, B2=17125579390, polynomial Dickson(12), sigma=0:12224056895737954441
dF=16384, k=6, d=158340, d2=11, i0=34
Expected number of curves to find a factor of n digits:
35 40 45 50 55 60 65 70 75 80
167 1024 7351 60402 558826 5744532 6.5e+07 7.8e+08 1.1e+10 2.8e+11
Step 1 took 67016ms
[/CODE]

vs.

[CODE]./yafu "ecm(2^1277-1,8)" -v -v -B1ecm 7000000


10/22/20 15:17:47 v1.35-beta @ cpu, System/Build Info:
Using GMP-ECM 7.0, Powered by GMP 6.2.0
detected Intel(R) Xeon(R) Gold 6248 CPU @ 2.50GHz
detected L1 = 32768 bytes, L2 = 28835840 bytes, CL = 64 bytes
measured cpu frequency ~= 2500.116400
using 1 random witnesses for Rabin-Miller PRP checks

===============================================================
======= Welcome to YAFU (Yet Another Factoring Utility) =======
======= bbuhrow@gmail.com =======
======= Type help at any time, or quit to quit =======
===============================================================
cached 664579 primes. pmax = 9999991


>> process id is 141455
commencing parallel ecm on 2601983048666099770481310081841021384653815561816676201329778087600902014918340074503059860433081046210605403488570251947845891562080866227034976651419330190731032377347305086443295837415395887618239855136922452802923419286887119716740625346109565072933087221327790207134604146257063901166556207972729700461767055550785130256674608872183239507219512717434046725178680177638925792182271
ECM has been configured with DIGITBITS = 52, VECLEN = 8, GMP_LIMB_BITS = 64
Choosing MAXBITS = 1456, NWORDS = 28, NBLOCKS = 7 based on input size 1277
linesieve took 0.022753 seconds
cached 5761455 primes < 99999989
Input has 1277 bits, using 1 threads (8 curves/thread)
Processing in batches of 100000000 primes
Initialization took 0.1013 seconds.

Building curves took 0.0007 seconds., B2=100*B1
commencing Stage 1 @ prime 2
Stage 1 took 219.2920 seconds
[/CODE]

So, 219.3/8 = 27.4 sec/curve versus 67 sec/curve.

AVX-ECM stage 2 took 118 sec, so 14.75 sec/curve, but these are just the standard continuation up to 100*B1.

It should scale linearly up to 7e9. But I've never run AVX-ECM with B1 anywhere close to that large. My bet is it crashes... but I'll test and see what happens.

storm5510 2020-10-22 23:48

[QUOTE=ryanp;560745][COLOR=Gray]More ECM on this number is not a waste, but it's becoming increasingly difficult. [/COLOR]:smile: And it's recommended that you or anyone else do curves at B1=76e8 or higher, maybe 1e10 or 2e10 now.[/QUOTE]

Thank you for the reply. You are the guy at the top of the ECM producers table. :grin:

It takes 155 minutes to run four curves in tandem at 16e8 on the machine where I have [I]Prime95[/I] running now. A single test with GMP-ECM, with no limit on B2, takes 50 minutes. 76e8, I would eventually get into this area. I could decrease my increment interval to one day, for example, until I got into higher areas. It is all about the time...

storm5510 2020-10-23 11:52

[U]Disregard my above...
[/U]
A while back, [B]VBCurtis[/B] suggested I start at 6e9. I have everything running to this level now. After a period of time, at least a month, I will go to 7e9, then continue with this progression in monthly intervals. I made a notation in my notebook with a red Sharpie marker. I would like to be able to run 500 tests with GMP-ECM on each group. Whether this many is possible, I will have to wait and see.


All times are UTC. The time now is 05:54.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.