[QUOTE=Prime95;510942]Please tell me you are using GMP-ECM for stage 2.[/QUOTE]
I plan to do so but not yet. Will I need to email you the results in this case? Or there is some way to submit it using manual testing results form?

[QUOTE=newalex;510953]I plan to do so but not yet. Will I need to email you the results in this case? Or there is some way to submit it using manual testing results form?[/QUOTE]

You will need to email me exponents, curve counts, and bounds. Later I can show you how to use the manual forms.

ok ,thanks.

P-1 found a factor in stage #2, B1=330000, B2=8745000, E=12.
UID: harlee/i5-5250U_1600, M18309659 has a factor: 324338066454357946195307829957109799 (P-1, B1=330000, B2=8745000, E=12)

117.965 bits

k= 2 × 43 × 353 × 1063 × 51713 × 131839 × 142421 × 565319

P-1 found a factor in stage #1, B1=685000.
UID: Jwb52z/Clay, M90890743 has a factor: 417698754453203763044685289 (P-1, B1=685000)

88.433 bits.

I have just started and found a factor on my 8 TF :D


M92700737 Factor: 11692583137317592272401 / TF: 73-74

[QUOTE=Thecmaster;511040]I have just started and found a factor on my 8 TF :D[/QUOTE]

Congratulations! And welcome. :smile:

Just so you know, you have been statistically "lucky" with this find. On average, you will find approximately one factor per 1/[bit level] where P-1'ing hasn't yet been done. Where P-1'ing has already been done (read: DCTF'ing) the success rate will be lower.

The reason I bring this up is please don't be discouraged if you don't find another factor for a while. Statistics has no memory, so every TF'ing run has the same probability of finding a factor, regardless of how "lucky" or "unlucky" you have been in the past. Run an infinite number of tests, and you will converge on the expected number of finds (assuming your kit is good).

P.S. Just to share, I personally use my (slow) GPUs to help Wayne's "to below 2,000" sub-sub-project. It took 426 (!) TF'ing runs from 71 to 72 bits to finally find [URL="https://www.mersenne.org/report_exponent/?exp_lo=33943387&full=1"]this factor[/URL], and get 33.9 to below 2,000 un-factored candidates.

[QUOTE=chalsall;511056]
P.S. Just to share, I personally use my (slow) GPUs to help Wayne's "to below 2,000" sub-sub-project. It took 426 (!) TF'ing runs from 71 to 72 bits to finally find [URL="https://www.mersenne.org/report_exponent/?exp_lo=33943387&full=1"]this factor[/URL], and get 33.9 to below 2,000 un-factored candidates.[/QUOTE]

All are welcome :)

[url]https://www.mersenneforum.org/showthread.php?t=22476[/url]

[QUOTE=chalsall;511056]On average, you will find approximately one factor per 1/[bit level][/QUOTE]
Are you sure? hihi :razz: That would be nice to have more factors than trials...

I though you find one factor in [bitlevel] tries.

Or my Englsih plays tricks to me? hihi

[QUOTE=LaurV;511194]
Or my [B]Englsih[/B] plays tricks to me? hihi[/QUOTE]

Your Englsih is certainly playing tricks on you... :smile: the English I don´t know.

[QUOTE=LaurV;511194]Are you sure? hihi :razz:[/QUOTE]

You're correct; I misspoke...

I should have said "your probability of finding a factor is approximately 1/[bit level], or one every [bit level] runs"....

M92617193 is divisible by 1175619298002469430597399521162713999664087. This 139.75-bit factor brought to you by P-1.

Some luck here. One more factor.

M55303189 has a factor: 18522127152677986447913 [TF:73:74:mfaktc 0.21 barrett76_mul32_gs]
found 1 factor for M55303189 from 2^73 to 2^74 [mfaktc 0.21 barrett76_mul32_gs]

UID: Jwb52z/Clay, M90939509 has a factor: 10728673174057307870137591 (P-1, B1=685000)

83.150 bits.

P-1 found a factor in stage #1, B1=685000.
UID: Jwb52z/Clay, M90969559 has a factor: 36906292085934351166387681 (P-1, B1=685000)

84.932 bits.

Today I found my largest composite factor 19147642464835832222111776488276027610060674573088897824886038321359 (223.5 bits - 68 digits) =82139596673583394582617549419848961 * 233110987151874257840851425095119 for the number of [URL="https://www.mersenne.org/report_exponent/?exp_lo=3146833&exp_hi=&full=1"]M3146833[/URL].

Impressive. Even the two component factors are quite respectable in their own rights (108, 116 bits).

[QUOTE=Miszka;512314]Today I found my largest composite factor 19147642464835832222111776488276027610060674573088897824886038321359 (223.5 bits - 68 digits) =82139596673583394582617549419848961 * 233110987151874257840851425095119 for the number of [URL="https://www.mersenne.org/report_exponent/?exp_lo=3146833&exp_hi=&full=1"]M3146833[/URL].[/QUOTE]

Can someone point me to a link of how to calculate the probability of P-1 finding a factor, as well as the GhZ-Day calculation for P-1? If I could figure out how to make a script to identify exponents < 40M which have similarly high likelihood of factors found/GhZ-Day, I might have to throw my home computers on them...

[url]https://www.mersenne.ca/prob.php[/url]
[url]https://www.mersenne.ca/p1small.php[/url] *

*this is actually pretty out-of-date and needs to be reworked, you may or may not find anything useful there.

157.7 Bits
 

2467763 Factored 300535914423293271697652047026544474483005087353

2019-03-15 Sid & Andy F-PM1 Factor: 300535914423293271697652047026544474483005087353 / (P-1, B1=6000000, B2=180000000, E=12)

[QUOTE=petrw1;512381]2467763 Factored 300535914423293271697652047026544474483005087353

2019-03-15 Sid & Andy F-PM1 Factor: 300535914423293271697652047026544474483005087353 / (P-1, B1=6000000, B2=180000000, E=12)[/QUOTE]
Wow, nice one, 158 bits, prime! Congrats!

P-1 found a factor in stage #1, B1=685000.
UID: Jwb52z/Clay, M91131653 has a factor: 169246433448198777407647199 (P-1, B1=685000)

87.129 bits.

320000-360000 range has been completed. 711 exponents were processed and 56 new factors were found. Now i am starting work on 360000-400000 range.


Also I found my first small P-1 factor
P-1 found a factor in stage #2, B1=100000000, B2=2000000000, E=12.
M287939 has a factor: 27265853414344474964957059537 (P-1, B1=100000000, B2=2000000000, E=12)

[QUOTE=newalex;512602]320000-360000 range has been completed. 711 exponents were processed and 56 new factors were found. Now i am starting work on 360000-400000 range.


Also I found my first small P-1 factor
P-1 found a factor in stage #2, B1=100000000, B2=2000000000, E=12.
M287939 has a factor: 27265853414344474964957059537 (P-1, B1=100000000, B2=2000000000, E=12)[/QUOTE]

Impressive

[QUOTE=newalex;510937](...) with a major goal to complete all unfactored <500000 to 30-digits level.[/QUOTE]

I´m gonna help you out with the remaining exponents in the 200k range.

Still, I´ll be much slower than you. Your rig is FAST, man!

yes, these two Xeon Scalable processors are awesome. Each core processes about 350 curves per day for B1=250k in this range.
I appreciate your help. Together we will make it done faster. I plan to work up to 500k first so 200k range is all your for 2-3 months.

My Kaby Lake @3GHz completes ~285/day per core for the same range.

Will be a bit more as soon as the 200k starts. 2 - 3 months will certainly be enough to complete the range.

The second P-1 factor for small exponents
P-1 found a factor in stage #2, B1=100000000, B2=2000000000, E=12.
M237959 has a factor: 8069793616575708974657372316228989319881 (P-1, B1=100000000, B2=2000000000, E=12)
132.568 bits

Nice find. 133 bits, and prime.
On another note, has anybody spotted the monster (first) factor recently found for M2357?
That was something: over 200 bits and very rough. Completely out of reach for P-1, as per the ludicrous number of ghz-days indicated in James' site.

So the only solution would be an ecm of a gnfs factorisation?

[url]https://www.mersenneforum.org/showthread.php?p=512799#post512799[/url]

P-1 found a factor in stage #1, B1=685000.
UID: Jwb52z/Clay, M90918851 has a factor: 4543028685782507004545407 (P-1, B1=685000)

81.910 bits.

[QUOTE=axn;512850][url]https://www.mersenneforum.org/showthread.php?p=512799#post512799[/url][/QUOTE]

[url]https://www.mersenneforum.org/showpost.php?p=513034&postcount=460[/url]

Thank you petrw1

[QUOTE=petrw1;513135][url]https://www.mersenneforum.org/showpost.php?p=513034&postcount=460[/url][/QUOTE]

Not exactly. M2357 is fully factored, as the remaining cofactor was certified prime, whereas M2557 is not yet at that stage for the cofactor is not a PRP.

But if you were referring to the much praised first factors for <10k exponents, then yes, lightning has striken twice... :beer:!

Wow a 3rd for Ryan
 

[QUOTE=lycorn;513285]But if you were referring to the much praised first factors for <10k exponents, then yes, lightning has striken twice... :beer:![/QUOTE]

2377 Factored 227428848043888440045045261533777705948742974440394358593

Date User Type Result
2019-04-13 Ryan Propper F-ECM Factor: 227428848043888440045045261533777705948742974440394358593

We may say lightning strikes thrice! :party:

I would like to know what program Ryan Propper is using. Although the results are reported as "F-ECM", I´m not so sure about it.

[QUOTE=lycorn;513634]I would like to know what program Ryan Propper is using. Although the results are reported as "F-ECM", I´m not so sure about it.[/QUOTE]

The three factors found so far in the 2k range are all impossibly non-smooth for P−1. The cofactors were surely impossibly large for NFS, they are about 650 digits even after the new factors.

Factors of 56, 57, and 62 digits are perfectly consistent with very deep but entirely feasible ECM, if you throw enough cores at it. Given Ryan's track record of envelope-pushing discoveries, he surely has sufficient resources at his disposal.

[QUOTE=GP2;513653]

Factors of 56, 57, and 62 digits are perfectly consistent with very deep but entirely feasible ECM, if you throw enough cores at it. Given Ryan's track record of envelope-pushing discoveries, he surely has sufficient resources at his disposal.[/QUOTE]

Yes, I agree. But my point has to do with the way the factors were reported.

First off, the Cleared Report has no mention to number of curves nor bounds used. Fair enough, there are other cases where we see that happen. User newalex, for one, who has recently been finding many factors for small exponents. He appears to be using AVX-512 AWS systems, that´s probably the reason.

Second, there are no traces of Ryan´s activity in the Exponent Status report for these exponents, except the finding.

Third, although I have no hard figures for that, I am under the impression that the number of curves recorded for these exponents hasn´t increased much lately. It could have been a case of great luck, of course, but, on [B]three[/B] exponents?

This reminds that when a factor was eventually found for M1061, a couple of years ago, F-ECM showed up in the report, although other method had been used. I think it was SNFS.

The importance of all this lies somewhere between nil and nothing, of course, but I just felt curious about it...:smile:

[QUOTE=lycorn;513666]Yes, I agree. But my point has to do with the way the factors were reported.[/QUOTE]

He is probably using GMP-ECM for stage 2. That would be the logical thing to do for such small exponents. And in that case there would be no automatic reporting to Primenet.

[QUOTE]This reminds that when a factor was eventually found for M1061, a couple of years ago, F-ECM showed up in the report, although other method had been used. I think it was SNFS.[/QUOTE]

But for M1061 it was split into two primes of 143 digits and 177 digits. With NFS you usually get enormous factors and the exponent ends up fully factored, and it's usually only feasible for the Cunningham ranges, in the low 1K exponent values. These factors and these exponents from Ryan Propper don't match that at all. They are much more typical of deep ECM.

Maybe he uses GMP-ECM and doesn't want to email results without factors? And uses a lot of computing power to compete 55-60 digits range for one exponent in few days.



I believe the problem with curves and bounds is the problem of version 29.6 which I am using. Perhaps it submits results in JSON format which is not parsed correctly by server. Text results in results.txt look identical to previous versions.


By the way I found another 40+ digits factor with 30 digits bounds. I am curious what is probability of such discovery?

[Sun Apr 14 07:19:46 2019]
ECM found a factor in curve #147, stage #2
Sigma=7835381084152872, B1=250000, B2=25000000.
M389749 has a factor: 35157083141838820140048845608443589051481 (ECM curve 147, B1=250000, B2=25000000)
134.69 bits

[QUOTE=GP2;513667]He is probably using GMP-ECM for stage 2. That would be the logical thing to do for such small exponents. And in that case there would be no automatic reporting to Primenet.[/QUOTE]

That makes sense. It was probably the case. It is just a pity the whole amount of curves done wasn"t reported.

woooohooooooo First P-1 factor

Factor: 73683712459653894665954956209427783 / (P-1, B1=695000, B2=13205000, E=12)

2477 makes 4 for Ryan
 

Beyond good luck...

Indeed. The default value (2500) for the ecm progress table will soon have to be raised or the table will display no values at all... :smile:

M2377 just had a factor found
 

This was the 6th smallest M without a factor!
[url]https://www.mersenne.org/report_ecm/[/url]
[url]https://www.mersenne.ca/exponent/2377[/url]

187 bits! and impossible to find with P-1 go ECM

P-1 found a factor in stage #1, B1=695000.
UID: Jwb52z/Clay, M91216309 has a factor: 646532530928667932245369 (P-1, B1=695000)

79.097 bits

[QUOTE=SethTro;513725]This was the 6th smallest M without a factor!
[url]https://www.mersenne.org/report_ecm/[/url]
[url]https://www.mersenne.ca/exponent/2377[/url]

187 bits! and impossible to find with P-1 go ECM[/QUOTE]Already reported just 10 posts before yours.

P-1 found a factor in stage #1, B1=695000.
UID: Jwb52z/Clay, M91221769 has a factor: 4115254730949418739220409 (P-1, B1=695000)

81.767 bits

90901697 F-PM1 2019-04-20 17:20 0.0 Factor: 462195198223850274269081 / (P-1, B1=695000)

Where and how can you check bitlevel on a factor?

[QUOTE=Thecmaster;514227]Where and how can you check bitlevel on a factor?[/QUOTE]Once the factor shows up on mersenne.ca (data sync happens nightly around 2am UTC) those details are available for each factor. If you want to quickly calculate the bitsize of any factor you can use the "Calculate bit size of factor" box at the bottom of the menu on any page of [url]www.mersenne.ca[/url] (78.613 bits in your case).

[QUOTE=James Heinrich;514229]Once the factor shows up on mersenne.ca (data sync happens nightly around 2am UTC) those details are available for each factor. If you want to quickly calculate the bitsize of any factor you can use the "Calculate bit size of factor" box at the bottom of the menu on any page of [url]www.mersenne.ca[/url] (78.613 bits in your case).[/QUOTE]

Oooh. Didn't know about that.

I found a factor a week ago: 73683712459653894665954956209427783 and that are 115.827 bits.

My largest factor so far. ;)

Thanx for info

[QUOTE=Thecmaster;514251]I found a factor a week ago: 73683712459653894665954956209427783 and that are 115.827 bits[/QUOTE]Not bad! I think I've only ever found 3 bigger than that, and I've been doing P-1 for a [i]long[/i] time :smile:
So then looking at [url]https://www.mersenne.ca/M90883987[/url] you can see the bitsize of the factor, as well as a graph showing how the factor fits into B1/B2 of the P-1, how it compares to TF that was done, etc.

edit: your factor is #485 on the list of largest P-1 factors ever found (out of 89,828).

[QUOTE=James Heinrich;514252]Not bad! I think I've only ever found 3 bigger than that, and I've been doing P-1 for a [i]long[/i] time :smile:
So then looking at [url]https://www.mersenne.ca/M90883987[/url] you can see the bitsize of the factor, as well as a graph showing how the factor fits into B1/B2 of the P-1, how it compares to TF that was done, etc.

edit: your factor is #485 on the list of largest P-1 factors ever found (out of 89,828).[/QUOTE]


Oooh. Is that so. I been doing this for a month and have been lucky so far. I found a factor my first day of TF. TF number 8 for me got a factor. But p-1 haven't yeald resault for me before this 115.827 bit factor. After 3 weeks of p-1 and after 4 weeks I now got an other factor.

Thanx for the stars if the factor.

Cheers and happy easter

[QUOTE=James Heinrich;514252]
edit: your factor is #485 on the list of largest P-1 factors ever found (out of 89,828).[/QUOTE]

Where did you find that list? 😁

I completed ECM for 360000-40000 range with 30-digits bounds.
36 new factors were found(one of them is 41 digits).

400000-440000 range is next.

[QUOTE=Thecmaster;514275]Where did you find that list? 😁[/QUOTE]Fiddling in my data. I've exposed a quick report that may be of general interest:
[url]https://www.mersenne.ca/pm1user[/url]

[QUOTE=James Heinrich;514321]Fiddling in my data. I've exposed a quick report that may be of general interest:
[url]https://www.mersenne.ca/pm1user[/url][/QUOTE]

Yeah. That's nice. And fun to check out. There are soon big factors in the top 100 factors. 😅

Ryan Propper has become the low-level Mersenne number killer.
He found the next factor, this time for M4111 - 10819968927585403636958816179849566933560096066787125831.
RPR is prime.
Congratulations!

[QUOTE=Miszka;515234]Ryan Propper has become the low-level Mersenne number killer.
He found the next factor, this time for M4111 - 10819968927585403636958816179849566933560096066787125831.
RPR is prime.
Congratulations![/QUOTE]

To find as many factors as he has should take Thousands of curves yet if you look at ECM top producers he is about 750th with .004 points and 8 factors in 8 attempts. Either he doesn't report curves run for unsuccessful attempts....or these aren't actually from ECM? OR????

A couple years ago I ran 1,000+ curves on each unfactored exponent under 20,000 for about 19,000 GhzDays. I found 2 factors under 9999.

Wouldn't the rest of us want to know how many curves are completed at each level so we know where best to continue to finish factoring them?

[QUOTE=petrw1;515241]Either he doesn't report curves run for unsuccessful attempts....[/QUOTE]

This is the simplest explanation

EDIT:- It is unlikely that he's using GIMPS/P95 to get assignments/report results. Most likely these are being run using GMP-ECM.

[QUOTE=axn;515246]
Most likely these are being run using GMP-ECM.[/QUOTE]

That sounds (very) reasonable. It does not explain why he is not getting any credit, even for the factors found, though.

[QUOTE=lycorn;515439]It does not explain why he is not getting any credit, even for the factors found, though.[/QUOTE]Maybe he doesn't care about credit on mersenne.org. I think the factors say enough.

P-1 found a factor in stage #1, B1=695000.
UID: Jwb52z/Clay, M91311949 has a factor: 1679230184991897832914127 (P-1, B1=695000)

80.474 bits

[QUOTE=Thecmaster;514227]90901697 F-PM1 2019-04-20 17:20 0.0 Factor: 462195198223850274269081 / (P-1, B1=695000)

Where and how can you check bitlevel on a factor?[/QUOTE]

I don't have the link at this second, but if you google "GIMPS Factor Bit Depth Calculator" there's a page that comes up where you click to activvate a right hand menu through 3 dark black lines at the top right of the screen and you past the factor into that bottom space and click off of it, but don't make the box go away and it will show you the bit number.

[QUOTE=Jwb52z;515462]I don't have the link at this second, but if you google "GIMPS Factor Bit Depth Calculator" there's a page that comes up where you click to activvate a right hand menu through 3 dark black lines at the top right of the screen and you past the factor into that bottom space and click off of it, but don't make the box go away and it will show you the bit number.[/QUOTE]That's literally any page on [url]www.mersenne.ca[/url], the bitsize calculator is at the bottom of the hamburger menu ≡.

[QUOTE=retina;515452]Maybe he doesn't care about credit on mersenne.org. I think the factors say enough.[/QUOTE]

It is perfectly possible, yes. But I would imagine that, independently of the user willingness for credit, the server would compute and display the corresponding amount whenever a completed work unit is received. This situation seems to indicate that, at least, the work done is not being submitted to the server in the same manner most users submit it.
And it is indeed a shame that the NF-ECM work (i.e. number of curves, if any) is not being recorded, for future reference.

[QUOTE=lycorn;515439]That sounds (very) reasonable. It does not explain why he is not getting any credit, even for the factors found, though.[/QUOTE]

I'm guessing this is because the factors are so large that they cause an overflow when the server tries to calculate how much credit to give. However, this wouldn't explain why there is a small amount of credit as opposed to no credit at all.

[QUOTE=Jwb52z;515462]I don't have the link at this second, but if you google "GIMPS Factor Bit Depth Calculator" there's a page that comes up where you click to activvate a right hand menu through 3 dark black lines at the top right of the screen and you past the factor into that bottom space and click off of it, but don't make the box go away and it will show you the bit number.[/QUOTE]Python can also be coaxed to do this for you locally without needing an Internet connection or a browser.[code]python -c "x= 10819968927585403636958816179849566933560096066787125831 ;import math;print(math.log(x,2))"[/code]Output:[code]182.819741575[/code]

[QUOTE=retina;515614]Python can also be coaxed to do this for you locally without needing an Internet connection or a browser.[code]python -c "x= 10819968927585403636958816179849566933560096066787125831 ;import math;print(math.log(x,2))"[/code]Output:[code]182.819741575[/code][/QUOTE]

or more puristic with bc (scale is for the number of decimal digits after the decimal point)

[code]
$ x=10819968927585403636958816179849566933560096066787125831; echo "scale=4; l($x)/l(2)" | bc -l
182.8320
[/code]

[QUOTE=vebis;515615]or more puristic with bc (scale is for the number of decimal digits after the decimal point)

[code]
$ x=10819968927585403636958816179849566933560096066787125831; echo "scale=4; l($x)/l(2)" | bc -l
182.8320
[/code][/QUOTE]That is less accurate.[code]x=6129982163463555433433388108601236734474956488734408704; echo "scale=4; l($x)/l(2)" | bc -l

182.0122[/code]The answer should be exactly 182.0000

[QUOTE=retina;515616]That is less accurate.[code]x=6129982163463555433433388108601236734474956488734408704; echo "scale=4; l($x)/l(2)" | bc -l

182.0122[/code]The answer should be exactly 182.0000[/QUOTE]

That's true, should have mentioned the scale is not only applied to the result.

[QUOTE=ixfd64;515567]I'm guessing this is because the factors are so large that they cause an overflow when the server tries to calculate how much credit to give. However, this wouldn't explain why there is a small amount of credit as opposed to no credit at all.[/QUOTE]

If PrimeNet is not told how many curves were run to find the factor is has assume only 1.
Further, I can't recall if Ryan's results give the bounds run. If not PrimeNet has to assume something. This might explain the low credit.

[CODE]
python3 -c "x= 10819968927585403636958816179849566933560096066787125831 ; print(x.bit_length())"

183
[/CODE]

Any calculation that uses log to determine bit size inevitably has rounding errors for very large numbers that are close to a power of two, where you get a result that is off by one. The [c]bit_length()[/c] result is always exact.

a new P43 factor was found in M3373 by... guess who?

If you guessed Ryan, you were wrong, it was James Hintz. The cofactor is still composite.

(You could have actually guessed James Hintz correctly, because he is responsible since the start of the year for factors that finished full factorization of M3203, M3089, M6679)

[QUOTE=DukeBG;515714]a new P43 factor was found in M3373 by... guess who?

If you guessed Ryan, you were wrong, it was James Hintz. The cofactor is still composite.[/QUOTE]

Ryan has only been finding factors of 55 digits or more. A mere 43-digit factor is like a penny on the ground that's not worth stooping to pick up. :smile:

Actually, Ryan seems to be focusing on unfactored exponents, which obviously already had pretty deep ECM done on them already. Whereas exponents like [M]M3373[/M] already had prior factors.

Ok, that's fair :grin:

[QUOTE=retina;515614]Python can also be coaxed to do this for you locally[/QUOTE]As can PHP:[code]> php -r "echo log(10819968927585403636958816179849566933560096066787125831, 2);"
182.81974157489[/code]

This is a personal best for me, to date:

[CODE]M83621 has a factor: 350415918423517661715474499722683369441 (ECM curve 1, B1=1000000, B2=100000000)[/CODE]128.042 bits.

:smile:

[QUOTE=storm5510;515818]This is a personal best for me, to date: 128.042 bits[/QUOTE]Not bad! I've only ever found 1 factor fractionally bigger ([SIZE="1"]128.605[/SIZE]), and I've never found an ECM factor. :down:

[QUOTE=James Heinrich;515837]Not bad! I've only ever found 1 factor fractionally bigger ([SIZE=1]128.605[/SIZE]), and I've never found an ECM factor. :down:[/QUOTE]

I have found only 12 out of 10,000+ ECM runs. Someone here once described this process as throwing darts at a dartboard. In this case, the dartboard is 50 feet away.

P-1 found a factor in stage #1, B1=700000.
UID: Jwb52z/Clay, M91556807 has a factor: 177424028897961775771913 (P-1, B1=700000),

77.232 bits.

After the latest factor found in the ranga 0 - 1M, exactly 80% of the numbers in that range are factored!
It was a 150-bit factor for M16369, found by... Ryan Propper.

First P-1 Factor
 

[M]M92129957[/M] has a factor: 538743045393521171980633 (78.834 bits)

This factor was found in P-1 Stage 1 with B1=1135000.

k = 2923821213731388 = 2^2 × 3^3 × 3187 × 8747 × 971149

The factor could have been found via trial factoring to 2^79. However, a fast and expensive GPU is necessary for efficiency, and 2^79 is past the ‘recommended’ trial factoring bound for M92129957.

Meanwhile, it appears that this link is broken: [URL="https://www.mersenne.ca/exponent/92129957"]https://www.mersenne.ca/exponent/92129957[/URL]

Please correct me if I made any mistakes.

[QUOTE=2M215856352p1;518086]Meanwhile, it appears that this link is broken: [URL="https://www.mersenne.ca/exponent/92129957"]https://www.mersenne.ca/exponent/92129957[/URL][/QUOTE]Apologies, it has been fixed now.

P-1 found a factor in stage #2, B1=700000, B2=12250000.
UID: Jwb52z/Clay, M91601753 has a factor: 58027651373023765151721470393 (P-1, B1=700000, B2=12250000),

95.551 bits.

M185223803 has a factor: 4094610292542423390791 (72 bits)

[QUOTE=lycorn;517984]After the latest factor found in the ranga 0 - 1M, exactly 80% of the numbers in that range are factored!
It was a 150-bit factor for M16369, found by... Ryan Propper.[/QUOTE]

It's been a while.....but he's back.

Ryan Propper Manual testing 7351 F-ECM 2019-06-22 03:02

[QUOTE=petrw1;519787]It's been a while.....but he's back.
Ryan Propper Manual testing 7351 F-ECM 2019-06-22 03:02[/QUOTE]You failed to mention it's a whopping 185.48 bit factor:[quote][M]M7351[/M] has a factor: [url=https://www.mersenne.ca/exponent/7351]68395426428631245237167716855062743466651615583157434009[/url][/quote]

Indeed. He´s been doing a tremendous job in the low ranges...
It´s just a pity we don´t know anything about his work. That is, if the amount of work he does [B]without[/B] finding factors was recorded, the picture of the ECM Progress table would certainly be quite different, and that could be used to better choose bounds while working in the low ranges. Oh well, nothing is perfect...:smile:

[QUOTE=James Heinrich;519839]You failed to mention it's a whopping 185.48 bit factor:[/QUOTE]

2.7x e^39 Tf days would have found it

P-1 found a factor in stage #1, B1=700000.
UID: Jwb52z/Clay, M91936463 has a factor: 756900670880501056009519 (P-1, B1=700000)

79.324 bits.

I've found 500 factors!
 

If anyone interested, I've just hit 500 factors.

Most of them is in the 900M range, from 2^70 to 2^71, I think the easiest way to find factors under 1G currently.

The 500th one is 1800724098004283562241, the factor of [URL="https://www.mersenne.org/report_exponent/?exp_lo=914797787&full=1"]M914797787[/URL]

Using six 3.8GHz cores of my i5-8400, I'm expecting to dig up 1000th factor by the end of 2019.

[QUOTE=Maciej Kmieciak;520053]If anyone interested, I've just hit 500 factors.

Most of them is in the 900M range, from 2^70 to 2^71, I think the easiest way to find factors under 1G currently.

The 500th one is 1800724098004283562241, the factor of [URL="https://www.mersenne.org/report_exponent/?exp_lo=914797787&full=1"]M914797787[/URL]

Using six 3.8GHz cores of my i5-8400, I'm expecting to dig up 1000th factor by the end of 2019.[/QUOTE]

Congratulations.... that's quite the commitment
If you really want lots more factors look into a GPU.
They are crazy fast at factoring

I was actually surprised to hear that anybody still uses CPU for TF in GIMPS range.

[QUOTE=Maciej Kmieciak;520053]If anyone interested, I've just hit 500 factors.

Most of them is in the 900M range, from 2^70 to 2^71, I think the easiest way to find factors under 1G currently.

The 500th one is 1800724098004283562241, the factor of [URL="https://www.mersenne.org/report_exponent/?exp_lo=914797787&full=1"]M914797787[/URL]

Using six 3.8GHz cores of my i5-8400, I'm expecting to dig up 1000th factor by the end of 2019.[/QUOTE]

Congratulations.... that's quite the commitment
If you really want lots more factors look into a GPU.
They are crazy fast at factoring

[QUOTE]I was actually surprised to hear that anybody still uses CPU for TF in GIMPS range.[/QUOTE]
[QUOTE]If you really want lots more factors look into a GPU.[/QUOTE]
I see but I don't know much about this kind of technology. I couldn't understand how to make the software work. :sad:

Should I use mfakto for AMD GPU? I downloaded it from [URL="https://github.com/Bdot42/mfakto"]here[/URL] and I found this in README file.
[QUOTE]#############################
# 1.2 Compilation (Windows) #
#############################

- Install AMD APP SDK >= 2.5
- Use the VS2010 solution to build the 32-bit or 64-bit binary, or
- use the Makefile to build using MinGW and gcc[/QUOTE]
What does it mean? Is there any place where I can learn how to do these steps?