mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Data > Marin's Mersenne-aries

Reply
 
Thread Tools
Old 2022-01-31, 13:14   #34
tServo
 
tServo's Avatar
 
"Marv"
May 2009
near the Tannhäuser Gate

19×41 Posts
Default

As requested, the last 2 proof files for 103,xxx,xxx and 108,xxx,xxx were uploaded over the weekend.
tServo is offline   Reply With Quote
Old 2022-02-01, 14:39   #35
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

11001010000002 Posts
Default

Only 4 remain available to claim for PRP with proof-generation, each of which will take a while:
PRP=1,2,276036017,-1,80,0 Kriesel did first PRP shifted type 1, available
PRP=1,2,315690521,-1,80,0 Kriesel did first PRP shifted type 1, available
PRP=1,2,369452123,-1,81,0 Kriesel did first PRP unshifted type 1, available
PRP=1,2,852348659,-1,86,0 Kriesel did first PRP unshifted type 1, available

Claimed but unable to reserve on mersenne.org:
PRP=1,2,231695603,-1,79,0 Kriesel did first PRP shifted type 1, Kriesel repeating with proof generation, ETA today
PRP=1,2,285754177,-1,80,0 Kriesel did first PRP shifted type 1, Kriesel repeating with proof generation, ETA 7 weeks

Last fiddled with by kriesel on 2022-02-01 at 14:41
kriesel is online now   Reply With Quote
Old 2022-02-01, 16:05   #36
S485122
 
S485122's Avatar
 
"Jacob"
Sep 2006
Brussels, Belgium

180510 Posts
Default

Quote:
Originally Posted by kriesel View Post
...
Claimed but unable to reserve on mersenne.org:
PRP=1,2,231695603,-1,79,0 Kriesel did first PRP shifted type 1, Kriesel repeating with proof generation, ETA today
PRP=1,2,285754177,-1,80,0 Kriesel did first PRP shifted type 1, Kriesel repeating with proof generation, ETA 7 weeks
I suppose the server will not assign them to you because you were the first time tester.
S485122 is online now   Reply With Quote
Old 2022-03-13, 16:10   #37
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

26·101 Posts
Default update

Only 2 remain available to claim for PRP with proof-generation, each of which will take a while:
PRP=1,2,315690521,-1,80,0 Kriesel did first PRP shifted type 1, available
PRP=1,2,852348659,-1,86,0 Kriesel did first PRP unshifted type 1, available

Claimed but unable to reserve on mersenne.org (because I ran the first PRP):
PRP=1,2,276036017,-1,80,0 Kriesel did first PRP shifted type 1, Kriesel repeating with proof generation, ETA ~16 days
PRP=1,2,285754177,-1,80,0 Kriesel did first PRP shifted type 1, Kriesel repeating with proof generation, ETA ~19 days
PRP=1,2,369452123,-1,81,0 Kriesel did first PRP unshifted type 1, Kriesel repeating with proof generation, ETA ~18 days
kriesel is online now   Reply With Quote
Old 2022-04-04, 22:06   #38
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

26×101 Posts
Default

Taking the last two, PRP with proof generation in Gpuowl:
PRP=1,2,315690521,-1,80,0 Kriesel did first PRP shifted type 1, ETA ~ 2 weeks
PRP=1,2,852348659,-1,86,0 Kriesel did first PRP unshifted type 1, ETA ~ 4 months
kriesel is online now   Reply With Quote
Old 2022-04-13, 16:34   #39
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

26·101 Posts
Default

Quote:
Originally Posted by kriesel View Post
Separately I am working on DC of the remaining Predict Mxx exponents. There are a lot of them, ~65 at the moment lacking DC.
There now remain 6 of those (one of which matches with a dubious claim made earlier). Combining the two, Predict Mxx lacking completed DC, and dubious claims lacking completed DC, but within range of feasible P-1 factoring, yields 7 exponents (for now). Current status of each follows:
258150181 (from predict M52 list) kriesel running PRP/proof, ETA ~ 30 April 2022
282362693 (from predict M52 list) kriesel running PRP/proof ETA ~ June 2022
315690521 (from dubious claims list) kriesel running PRP/proof ETA ~ 19 April 2022
----limit of mersenne.org current server hardware automatic handling of proof files is ~ 596M----
655685803 (from predict M45 list) kriesel running PRP/proof ETA July 2022
852348659 (from dubious claims and subsequently predict M52 list) kriesel running PRP/proof ETA July 2022
----limit of mersenne.org exponent range 1000M----
1000000007 (from predict M52 list) TF and P-1 done; available for first PRP/proof
----limit of mprime/prime95 AVX512 64M fft exponent handling (PRP, proof, cert,...) 1169M----
----limit of PRP/proof-capable gpuowl versions 2172.36M----
2252945257 (from predict M51 list) TF done, Mlucas P-1 stage 1 running, at ~23% completed now, total time for both stages estimated at 9.8 months on an Intel i5-7600T; no first PRP test yet, waiting on P-1 GCD results

~1G PRP/proof is feasible in ~4.5 months in Gpuowl on a Radeon VII.

If 2.253G were supported in some proof-capable gpuowl version, its PRP/proof would take ~ 2. years on a Radeon VII.
From a 268 msec/iteration timing on Mlucas in 128M fft, on Ubuntu/WSL/Win10, using 24 threads on all 24 real cores of a dual Xeon e5-2697v2 system, it would take ~19.1 years there. I think one could do quite a bit better than that in Mlucas on native Linux, & perhaps employing more threads and/or different hardware, but it is still likely to take at least several years to run on any CPUs I have. And currently, proof generation is not released in Mlucas, and Cert capability is not available, in any known software for such a high exponent, so 2 full PRP runs would be necessary. Proof generation on Mlucas may become available before the P-1 stages complete, or shortly after. That would leave unresolved how to perform a cert for an exponent nearly double the maximum now supported in mprime/prime95.

Last fiddled with by kriesel on 2022-04-13 at 16:38
kriesel is online now   Reply With Quote
Old 2022-05-14, 17:35   #40
kriesel
 
kriesel's Avatar
 
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest

26×101 Posts
Default

Here's a quick update edited into the quoted block (with both deletions and additions):
Quote:
Originally Posted by kriesel View Post
There now remain 6 4 of those (one of which matches with a dubious claim made earlier). Combining the two, Predict Mxx lacking completed DC, and dubious claims lacking completed DC, but within range of feasible P-1 factoring, yields 7 5 exponents (for now). Current status of each follows:
[M]258150181[/M] (from predict M52 list) kriesel running PRP/proof, ETA ~ 30 April 2022 done
282362693 (from predict M52 list) kriesel running PRP/proof ETA ~ June 25 2022
[M]315690521[/M] (from dubious claims list) kriesel running PRP/proof ETA ~ 19 April 2022 done
----limit of mersenne.org current server hardware automatic handling of proof files is ~ 596M----
655685803 (from predict M45 list) kriesel running PRP/proof ETA July 2022
852348659 (from dubious claims and subsequently predict M52 list) kriesel running PRP/proof ETA July ~22 2022
----limit of mersenne.org exponent range 1000M----
1000000007 (from predict M52 list) TF and P-1 done; available for first PRP/proof
----limit of mprime/prime95 AVX512 64M fft exponent handling (PRP, proof, cert,...) 1169M----
----limit of PRP/proof-capable gpuowl versions 2172.36M----
2252945257 (from predict M51 list) TF done, Mlucas P-1 stage 1 running, at ~23 52% completed now, total time for both stages estimated at 9.8 months on an Intel i5-7600T; no first PRP test yet, waiting on P-1 GCD results, then wait for Mlucas V21 proof capable version release (but run time would be very long on any hardware I have)

~1G PRP/proof is feasible in ~4.5 months in Gpuowl on a Radeon VII.

If 2.253G were supported in some proof-capable gpuowl version, its PRP/proof would take ~ 2. years on a Radeon VII.
From a 268 msec/iteration timing on Mlucas in 128M fft, on Ubuntu/WSL/Win10, using 24 threads on all 24 real cores of a dual Xeon e5-2697v2 system, it would take ~19.1 years there. I think one could do quite a bit better than that in Mlucas on native Linux, & perhaps employing more threads and/or different hardware, but it is still likely to take at least several years to run on any CPUs I have. And currently, proof generation is not released in Mlucas, and Cert capability is not available, in any known software for such a high exponent, so 2 full PRP runs would be necessary. Proof generation on Mlucas may become available before the P-1 stages complete, or shortly after. That would leave unresolved how to perform a cert for an exponent nearly double the maximum now supported in mprime/prime95.

Last fiddled with by kriesel on 2022-05-14 at 17:39
kriesel is online now   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
"Mixed Sierpinski conjecture base 5" proven!! (if probable primes allowed) sweety439 Sierpinski/Riesel Base 5 1 2020-12-21 14:08
Is "mung" or "munged" a negative word in a moral sense? Uncwilly Lounge 15 2020-04-14 18:35
Aouessare-El Haddouchi-Essaaidi "test": "if Mp has no factor, it is prime!" wildrabbitt Miscellaneous Math 11 2015-03-06 08:17
Palindrome primes (a.k.a. Elementary S03E03 "Just a Regular Irregular") Batalov And now for something completely different 12 2014-11-16 19:03
Would Minimizing "iterations between results file" may reveal "is not prime" earlier? nitai1999 Software 7 2004-08-26 18:12

All times are UTC. The time now is 14:58.


Mon May 16 14:58:01 UTC 2022 up 32 days, 12:59, 1 user, load averages: 2.11, 1.65, 1.49

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔