332.2M - 333.9M (aka 100M digit range) - Page 44

kracker · 2013-10-20, 00:37

I agree with what Uncwilly says, on "Do what you like best". There are valid reasons why someone would want to test a 100M(332M) exponent.

(this excludes poaching, obviously)

Uncwilly · 2014-02-21, 05:34

Here is the overdue status report for the range from 332192831 to 332399999 (plus other data out to 332999999):

Code:

Date of data				2014/02/20
Average bit depth for first 100 expos	78.57
Average bit depth for first 1000 expos	77.81
Average bit depth for first 5000 expos	75.05
Average bit depth for first 10000 expos	74.04
100th active expo (no factor found)	332198423
1000th active expo (no factor found)	332246549
5000th active expo (no factor found)	332443807
10000th active expo (no factor found)	332685701
Unitless total effort number		291,172,352
Number of first 100 expos to 2^77	90
Number of first 1000 expos to 2^76	992
Number of first 5000 expos to 2^77	1108
Number of first 10000 expos to 2^76	1889
Number left in range			4124

Estimated expos in range to be removed	188
(by taking all expos to 2^79)
Estimated expos in range to be removed	332
(by taking all expos to 2^82)

Code:

Bit	# at bit level
73	1107
74	560
75	673
76	713
77	287
78	66
79	680
80	26 (had been all the way up to 27)
81	11
82	0
83	0
84	0
85	1
P-1     332

The attached shows the total effort, by bit level for the range from 332192831 to 332999959

Uncwilly · 2014-05-21, 05:25

Here is the overdue status report for the range from 332192831 to 332399999 (plus other data out to 332999999):

Code:

Date of data				2014/05/20
Average bit depth for first 100 expos	78.72
Average bit depth for first 1000 expos	77.85
Average bit depth for first 5000 expos	75.21
Average bit depth for first 10000 expos	74.13
100th active expo (no factor found)	332198423
1000th active expo (no factor found)	332246639
5000th active expo (no factor found)	332444711
10000th active expo (no factor found)	332686891
Unitless total effort number		300,220,416
Number of first 100 expos to 2^77	100
Number of first 1000 expos to 2^76	1000
Number of first 5000 expos to 2^77	1135
Number of first 10000 expos to 2^76	2064
Number left in range			4106

Estimated expos in range to be removed	177
(by taking all expos to 2^79)
Estimated expos in range to be removed	320
(by taking all expos to 2^82)

Code:

Bit	# at bit level
73	926
74	267
75	955
76	860
77	311
78	67
79	680
80	28
81	11
82	0
83	0
84	0
85	1
P-1     343

The 332M range is currently has 59.83% of the exponents factored out.

Uncwilly · 2014-09-22, 04:33

I am still plunking away at this range.
Here is the status report for the range from 332192831 to 332399999 (plus other data out to 332999999):

Code:

Date of data				2014/09/21
Average bit depth for first 100 expos	78.73
Average bit depth for first 1000 expos	77.87
Average bit depth for first 5000 expos	75.31
Average bit depth for first 10000 expos	74.17
100th active expo (no factor found)	332198423
1000th active expo (no factor found)	332246639
5000th active expo (no factor found)	332198423
10000th active expo (no factor found)	332246639
Unitless total effort number		310,251,520
Number of first 100 expos to 2^78	81
Number of first 1000 expos to 2^77	734
Number of first 5000 expos to 2^77	1311
Number of first 10000 expos to 2^76	2235
Number left in range			4096

Estimated expos in range to be removed	170
(by taking all expos to 2^79)
Estimated expos in range to be removed	313
(by taking all expos to 2^82)

Code:

Bit	# at bit level
73	841
74	308
75	818
76	857
77	480
78	71
79	679
80	30
81	11
82	0
83	0
84	0
85	1
P-1     364

The 332M range is currently has 59.85% of the exponents factored out.

I have lost 4 cores and may lose another 4 soon (Borgs that were idle, but may actually be employed).

Uncwilly · 2014-12-21, 05:03

Here is the status report for the range from 332192831 to 332399999 (plus other data out to 332999999):

Code:

Date of data				2014/12/20
Average bit depth for first 100 expos	78.75
Average bit depth for first 1000 expos	77.89
Average bit depth for first 5000 expos	75.41
Average bit depth for first 10000 expos	74.22
100th active expo (no factor found)	332198423
1000th active expo (no factor found)	332246639
5000th active expo (no factor found)	332445497
10000th active expo (no factor found)	332687423
Unitless total effort number		315,195,392
Number of first 100 expos to 2^78	83
Number of first 1000 expos to 2^77	756
Number of first 5000 expos to 2^77	1356
Number of first 10000 expos to 2^76	2318
Number left in range			4090

Estimated expos in range to be removed	164
(by taking all expos to 2^79)
Estimated expos in range to be removed	307
(by taking all expos to 2^82)

Code:

Bit	# at bit level
73	592
74	426
75	863
76	893
77	521
78	74
79	679
80	30
81	11
82	0
83	0
84	0
85	1
P-1     379

For the following bit levels, these are the 'hold-up' numbers (they are the lowest exponent that is factored to less than the indicated bit level.
80 - 332192831
79 - 332192969
78 - 332193937
77 - 332205547
76 - 332259079
75 - 332303119
74 - 332356067
73 - 332751583
We have been working of late on moving some of these forward.
The attached graph indicates the progress since last report.
TJAOI has reported a number of factors in this range, for numbers that already had known factors.

Uncwilly · 2015-02-21, 19:56

Here is the status report for the range from 332192831 to 332399999 (plus other data out to 332999999):

Code:

Date of data				2015/02/20
Average bit depth for first 100 expos	78.88
Average bit depth for first 1000 expos	77.91
Average bit depth for first 5000 expos	75.78
Average bit depth for first 10000 expos	74.42
100th active expo (no factor found)	332198423
1000th active expo (no factor found)	332246639
5000th active expo (no factor found)	332446357
10000th active expo (no factor found)	332688557
Unitless total effort number		324,354,048
Number of first 100 expos to 2^78	95
Number of first 1000 expos to 2^77	756
Number of first 5000 expos to 2^77	1372
Number of first 10000 expos to 2^76	2777
Number left in range			4082

Estimated expos in range to be removed	155
(by taking all expos to 2^79)
Estimated expos in range to be removed	298
(by taking all expos to 2^82)

Code:

Bit	# at bit level
73	494
74	291
75	812
76	1154
77	520
78	89
79	677
80	33
81	11
82	0
83	0
84	0
85	1
P-1     401

For the following bit levels, these are the 'hold-up' numbers (they are the lowest exponent that is factored to less than the indicated bit level.
80 - 332192831
79 - 332192969
78 - 332197721
77 - 332205547
76 - 332264497
75 - 332316329
74 - 332356991
73 - 332785007
We have been working of late on moving some of these forward. Current plans are to see all of the first 100 to 78, then moving that CPU to a different set of assignments.

aurashift · 2015-04-24, 05:00

So um... How should I go about getting exponents that have been factored enough to begin LL immediately? Or what should I be doing?

LaurV · 2015-04-24, 05:58

You click here. Remark the selection of checkbox and the bitlevels. The list is (usually) pretty short. You can play around (like changing the upper limit, bitlevels, or unchecking the "reserved" box to see the difference). As the LL in this range takes months, you can do this manually from time to time. The best/fast way is still to select any expo you like [i.e. not in the list] and ask on the forum for somebody to factor it. Don't be shy, we are here to help

Now, endless discussion can be done about the fact that the most of the reserved exponents will never be finished, most "beginners" come and reserve an exponent because they read something in a news paper or web page, and they expect to find a 100M decimal digits prime by tomorrow and go home with $50k from the EFF's money. When they see how long it takes, they just give up, without unreserving, and never return. As the exponents in this range are not (so often) recycled, the reservations persist for a long time.

So, if you really want an expo which is reserved, because you dreamed last night that it may be prime, just say so. Someone will check if the reservation shows any progress, and most probably you can have it, if you are really willing to invest that HUGE computing power you have, into it.

@Madpoo, as we see you like to play with that database

, maybe this is a nice homework, to make a list of 100M dec digits exponents which are reserved for LL by the people who never did (or did too less) LL work in lower ranges.

Uncwilly · 2015-04-24, 07:00

Some of the exponents that have a good amount of TF may still need P-1.

This (link) will produce a list of available exponents that meet a reasonable level of TF and are available. The ideal bit level is 81 if there is a GPU to do the work, else it is 78.

If you have a bunch of memory available, check with this tool of James's to make sure that the exponent has had enough P-1 for about 4-5% probability of finding a factor, before you take the number for LL.

aurashift · 2015-04-29, 15:56

Would it be possible to unreserve ~60 abandoned exponents shortly before I get those ~30 blades I'm expecting in?

lycorn · 2015-04-29, 19:20

Quote:

Originally Posted by aurashift

(...) those ~30 blades I'm expecting in?

May I at least try to lure you into giving a hand to this small sub-project (GMP-ECM on very small numbers, namely the (in)famous M1277)? The idea of using all that memory is most enticing...

http://www.mersenneforum.org/showthread.php?t=20092

2015-04-24, 05:58	#481
LaurV Romulan Interpreter Jun 2011 Thailand 22677₈ Posts	You click here. Remark the selection of checkbox and the bitlevels. The list is (usually) pretty short. You can play around (like changing the upper limit, bitlevels, or unchecking the "reserved" box to see the difference). As the LL in this range takes months, you can do this manually from time to time. The best/fast way is still to select any expo you like [i.e. not in the list] and ask on the forum for somebody to factor it. Don't be shy, we are here to help Now, endless discussion can be done about the fact that the most of the reserved exponents will never be finished, most "beginners" come and reserve an exponent because they read something in a news paper or web page, and they expect to find a 100M decimal digits prime by tomorrow and go home with $50k from the EFF's money. When they see how long it takes, they just give up, without unreserving, and never return. As the exponents in this range are not (so often) recycled, the reservations persist for a long time. So, if you really want an expo which is reserved, because you dreamed last night that it may be prime, just say so. Someone will check if the reservation shows any progress, and most probably you can have it, if you are really willing to invest that HUGE computing power you have, into it. @Madpoo, as we see you like to play with that database , maybe this is a nice homework, to make a list of 100M dec digits exponents which are reserved for LL by the people who never did (or did too less) LL work in lower ranges. Last fiddled with by LaurV on 2015-04-24 at 06:09

2015-04-29, 15:56	#483
aurashift Jan 2015 11×23 Posts	Would it be possible to unreserve ~60 abandoned exponents shortly before I get those ~30 blades I'm expecting in? Last fiddled with by aurashift on 2015-04-29 at 15:56

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
GPU72 / MISFIT use for 100M digit range?	Uncwilly	GPU to 72	64	2013-03-31 02:45
I want a 100M digit Mersenne that....	JuanTutors	PrimeNet	8	2012-12-06 13:47
How far along are you in your 100M digit LL test?	JuanTutors	Lounge	6	2012-02-21 07:36
100M-digit n/k pairs	__HRB__	Riesel Prime Search	0	2010-05-22 01:17
100M digit prime	Unregistered	Information & Answers	10	2010-03-24 20:16

2013-10-20, 00:37	#474
kracker "Mr. Meeseeks" Jan 2012 California, USA 2³·271 Posts	I agree with what Uncwilly says, on "Do what you like best". There are valid reasons why someone would want to test a 100M(332M) exponent. (this excludes poaching, obviously)

2015-04-24, 05:00	#480
aurashift Jan 2015 11·23 Posts	So um... How should I go about getting exponents that have been factored enough to begin LL immediately? Or what should I be doing?

2015-04-24, 07:00	#482
Uncwilly 6809 > 6502 """"""""""""""""""" Aug 2003 101×103 Posts 10011001110010₂ Posts	Some of the exponents that have a good amount of TF may still need P-1. This (link) will produce a list of available exponents that meet a reasonable level of TF and are available. The ideal bit level is 81 if there is a GPU to do the work, else it is 78. If you have a bunch of memory available, check with this tool of James's to make sure that the exponent has had enough P-1 for about 4-5% probability of finding a factor, before you take the number for LL.