User TJAOI

[Mark Rose]

Originally Posted by LaurV

Many of them were multiple bits (like 65 to 67, and I even have seen one like 63 to 68? So if you split them, you had much more than 1200 assignments).

There were approximately 1.89 assignments per exponent in the big list, so for 1 in 31, that's about 1 in 58 per exponent bit level. That's not far off.

[TheMawn]

Originally Posted by axn

Fun exercise: At what error rate is it worthwhile to redo factoring 64-68 range instead of freshly doing 74-75 on the same exponent?

This is of course a very important thing to keep in mind. I'm just seeing a few people starting to find three or four misses for one user and go blindly repeating work and I'm trying to show that this isn't always the best course of action.

The answer to the question is 0.01% or greater at 64 bits. It's way up to 1.5% at 68 bits. This is what I was hinting at when I said that it doesn't take long for the work to become less worthwhile at higher bit levels.

Originally Posted by bloodIce

OK, this holds, but my question is: How do you know the error rate a priori? How do you know it is 1 to 4, 1 to 2 or 1 to 10?
On non-compromised machine the error rate is 1 to trillion or less, but how much will be on a faulty hardware? How faulty? How do you know?

And second question: Why do you think that everyone should do something that is the most optimal? Where is the fun factor. At the end GIMPS is a voluntary project, not a military aim. Why not let me take the risk of a TF below the expected probability for a factor? That is close to take 100 random TF assignments form PrimeNet and to believe that at least one factor will appear. Obviously not guaranteed. So, should you do it at all?

The answer to your first question is strictly about probability. Over a long enough time, one expects to find a certain number of factors after however many attempts. If one is significantly below the threshold, the idea is that one has missed a bunch of factors. Back with my example, if I am expected to find 1 in 100, then after 10,000 I should have 100 factors. But, if I have only found 75, then the probabilities suggest that I have missed 25.

Missing 25 out of 100 --> 1 in 4 miss rate. Still, this is only valid on a per-hardware basis because as we saw, only sannerud.com's laptop was having trouble, so his account stats overall might have looked okay.


Regarding the second question, note that my explanation began with "The goal of this project is to find the next Mersenne Prime". It is completely fine for you to have your own goal, in which case my math means nothing because it began with a false assumption.

Some people have been commenting on "amount of saved work" including you. If your goal is to find factors as opposed to most efficiently finding the next Mersenne Prime, then you can stop talking about the "amount of saved work" because you don't care about the "amount of saved work". You care about the factors.

Someone who does care about the "amount of saved work" should just make sure that the cost of saving the LL tests is not more than performing the LL tests themselves.

[Mark Rose]

I ended up with 59 exponents out of 5770 attempts, or about 1.02%. Not that bad.

But 21 were already factored by someone else at the same TF bit level, at a higher TF bit level, through P-1 factoring, or the unknown mechanism through which TheJudger reported a few factors.

MM51890519 has a factor: 422835347805002599631 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52138049 has a factor: 535974705886413199687 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52562327 has a factor: 296938917474700320121 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52722451 has a factor: 373035884518340430911 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52733041 has a factor: 525148122526114916273 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] P-1 only
MM52736731 has a factor: 467377202976495825521 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52744631 has a factor: 441418946165759376247 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52752289 has a factor: 386313131512887140663 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52761881 has a factor: 515978965975485900689 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52762363 has a factor: 575234307694178809343 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52777031 has a factor: 418383038793943568999 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM52780037 has a factor: 533917432484655670393 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM52799741 has a factor: 471451500830029816913 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Factored at higher TF
MM52806697 has a factor: 417940079224832809321 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM53382619 has a factor: 311865198662999886761 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] DC
MM56921869 has a factor: 280913622489522427247 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM60003079 has a factor: 883300283732743033433 [TF:69:70*:mfaktc 0.20 barrett76_mul32_gs] Assigned DC?
MM62826661 has a factor: 65418875440186239329 [TF:65:67*:mfaktc 0.20 barrett76_mul32_gs] Factored at higher TF
MM63668599 has a factor: 291324354417694843079 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] LL
MM63730087 has a factor: 208953409763778958711 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Factored at higher TF
MM64254121 has a factor: 365797987313559100369 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] TF'ed by someone else
MM64548101 has a factor: 521585933979200069039 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Assigned DC?
MM65692787 has a factor: 157451764492341208247 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM66558353 has a factor: 310210900340264976113 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Different factor found by P-1
MM66646717 has a factor: 74655144016241015911 [TF:66:67*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM66656749 has a factor: 355055780099422618529 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM66826681 has a factor: 182880165473838614839 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by ??
MM66850621 has a factor: 54014626454123861087 [TF:65:67*:mfaktc 0.20 barrett76_mul32_gs] Different factor found by P-1
MM66991783 has a factor: 82619526561787732367 [TF:66:67*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM67214051 has a factor: 332298963464640961039 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Assigned DC?
MM67896701 has a factor: 570046652926508214199 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Assigned DC?
MM68075017 has a factor: 472834979311744538807 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by ??
MM68336699 has a factor: 106263897967969956001 [TF:64:67*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68343427 has a factor: 44081958540117101921 [TF:64:67*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68365613 has a factor: 258779248069492283873 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68389187 has a factor: 52700763107122717343 [TF:65:66*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68389943 has a factor: 459460649758956277721 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by ??
MM68470657 has a factor: 65789962404747450313 [TF:64:67*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68621681 has a factor: 320729348220772757473 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM68660923 has a factor: 188399229397513595239 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Possibly assigned LL and/or P-1
MM68723147 has a factor: 269124663607229949599 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM68751467 has a factor: 272933318409841166999 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68781347 has a factor: 112647003103135552151 [TF:65:67*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM68819329 has a factor: 372086203916326209001 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68850539 has a factor: 273503990877618318281 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68949029 has a factor: 362997221262109799719 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68978089 has a factor: 193423066443318976711 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM68982461 has a factor: 87962659386756805559 [TF:66:67*:mfaktc 0.20 barrett76_mul32_gs] LL
MM69023419 has a factor: 196233427174962269089 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] LL
MM69053969 has a factor: 575682621371008960247 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM69117571 has a factor: 531806295266636553593 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM69160837 has a factor: 389877077449592894911 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] LL
MM69410291 has a factor: 95521634531194089737 [TF:66:67*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by P-1
MM69419309 has a factor: 575651634606598099129 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM69625627 has a factor: 286151817309697609481 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs] Assigned LL?
MM69655601 has a factor: 143432579467186073513 [TF:65:67*:mfaktc 0.20 barrett76_mul32_gs] LL
MM69688097 has a factor: 478475433180285062609 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs] Different factor found by TF
MM69858031 has a factor: 22636497128783359687 [TF:64:66*:mfaktc 0.20 barrett76_mul32_gs] Same factor found by ??
MM80072843 has a factor: 50231130793387599631 [TF:65:66*:mfaktc 0.20 barrett76_mul32_gs] TF'ed by someone else

All the previous assignments share the same expiry when a factor is found, so I have no way of knowing if LL and DC were assigned at the time.

So it looks like I saved some DC work by doing the below 70M stuff. I'll let someone smarter than me determine whether it was worth the effort or not.

[tigreroars]

Originally Posted by tigreroars

So Far

M73067893 has a factor: 500353788552269706049
M73196477 has a factor: 405689354724536129129
M73200653 has a factor: 222147635215583971639
M73232539 has a factor: 97353251934198098303
M73250131 has a factor: 32954849667884776097
M73319819 has a factor: 127504245927620817289
M73339793 has a factor: 43784679984484989647

M73411021 has a factor: 516273780573731030777
M73447019 has a factor: 91937112001062099839
M73472723 has a factor: 237563045570961322129
M73508353 has a factor: 42519070002438362383
M73519793 has a factor: 42787351569247662031
M73582363 has a factor: 87074482214913222241
M73624877 has a factor: 141480500796788850239


Still crunching, ETA stands at about 2 more days (eye balling it, don't have those fancy calculators or computers you guys have for judging how much I have left )

[petrw1]

Originally Posted by tigreroars

(eye balling it, don't have those fancy calculators or computers you guys have for judging how much I have left )

Are you using MisFit?
It will tell you on the stats window; that's all I have to know how long it should take.

[tigreroars]

Originally Posted by petrw1

Are you using MisFit?
It will tell you on the stats window; that's all I have to know how long it should take.

don't have those fancy calculators or computers you guys have for judging how much I have left

Txt documents ought to be enough for anybody.

[kladner]

2 more while I was at work. 50 to go. Should finish sometime this morning.

M79587467 has a factor: 310235208858106146313 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs]
M79489517 has a factor: 145004329106888022433 [TF:66:67*:mfaktc 0.20 barrett76_mul32_gs]

M79483499 has a factor: 262360609558645869361 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs]
M79466503 has a factor: 158171873142735543391 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs]
M79406497 has a factor: 142926079712600712041 [TF:66:67*:mfaktc 0.20 barrett76_mul32_gs]

M79359781 has a factor: 86018707765317281777 [TF:66:67*:mfaktc 0.20 barrett76_mul32_gs]
M79329049 has a factor: 436594536103628352239 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs]
M79227053 has a factor: 465022048023243077879 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs]
M79004249 has a factor: 234687434795636918167 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs]
M79033063 has a factor: 463838510863801116457 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs]

M79388209 has a factor: 635545254030291494921 [TF:69:70*:mfaktc 0.20 barrett76_mul32_gs]
M79387769 has a factor: 232488207444013028009 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs]
M79383181 has a factor: 278890815675960387673 [TF:67:68*:mfaktc 0.20 barrett76_mul32_gs]
M79373999 has a factor: 42176261588351110289 [TF:65:67*:mfaktc 0.20 barrett76_mul32_gs]
M79362671 has a factor: 396626201607035962649 [TF:68:69*:mfaktc 0.20 barrett76_mul32_gs]

[bloodIce]

Originally Posted by TheMawn

Regarding the second question, note that my explanation began with "The goal of this project is to find the next Mersenne Prime". It is completely fine for you to have your own goal, in which case my math means nothing because it began with a false assumption.

Some people have been commenting on "amount of saved work" including you. If your goal is to find factors as opposed to most efficiently finding the next Mersenne Prime, then you can stop talking about the "amount of saved work" because you don't care about the "amount of saved work". You care about the factors.

Someone who does care about the "amount of saved work" should just make sure that the cost of saving the LL tests is not more than performing the LL tests themselves.

I can accept your point of view in the scope of the whole project. However, for my case the situation is different. I am able to find factors of these missed exponents in 58M range (1 factor per 2 days of work). So lets say so far I clear an exponent every second day in that range (with a factor). To do a LL-test on my GPU of a single one will take me about 2 months+. On the CPU the time is considerably more. So, by TFing an error list I am contributing more efficiently to the project, than doing even a single LL test. And yes, if error rate is higher, I will be even more efficient. If George points me to a machine ... my success rate would have been sky-high. But no, I am just re-testing couple of thousands expos with a spider. And still that makes sense. At least to me.

P.S.: I saved two or three LL test (depending on the point of view, since I aborted one). That is saved work, no matter how you score it. That is my whole point of this double-check. I am not advocating a double check on the whole range, but on one user who has documented errors. That is a big difference.

P.S.: Higher bit levels (TF74 and TF75) will have very little efficiency in the same range if there is no luck. Pminus1 is not finding all factors even to TF65 as we see and still so far is only a CPU exercise.

[LaurV]

Originally Posted by bloodIce

So lets say so far ...

+1. That is what I was talking about.

[kladner]

79-80 finished. No more factors.

[tigreroars]

73M to 74M checked, 21 F total.