BOINC NFS sieving - NFS@Home - Page 118

fivemack · 2014-04-14, 12:56

Carlos: Thanks for your offer, but there's no need. Let the server idle for a while.

fivemack · 2014-04-14, 15:00

I'm being inconsistent with myself.

One thing that I've become really sure of, in doing quite large numbers of factorisations with moderately significant resources, is that the circumstances in which it's worth doing more sieving to make the matrix smaller are rare. Particularly rare if you regard watt-hours as the currency to work in, and even more particularly rare if you regard third parties' watt-hours as worth more than yours.

For example, for 2+947, I spent a month on 56 threads, 700W of equipment in order to make the linear algebra take two weeks less using 12 threads on 200W equipment.

So, if you haven't already pushed more jobs for L1247, don't. It'll take longer even in wall-clock than the time saved, and much longer in terms of coal burned (or, I suppose, water through the Jirau dam). I'll run with the matrix as it is.

frmky · 2014-04-14, 19:04

Quote:

Originally Posted by fivemack

So, if you haven't already pushed more jobs for L1247, don't. It'll take longer even in wall-clock than the time saved, and much longer in terms of coal burned (or, I suppose, water through the Jirau dam). I'll run with the matrix as it is.

Ok, I'll leave it as it is. Thanks!

pinhodecarlos · 2014-04-14, 20:04

The offer still stands...I can start it tomorrow at this time. Please let me know.

fivemack · 2014-04-14, 22:31

Go ahead

pinhodecarlos · 2014-04-15, 01:24

Quote:

Originally Posted by fivemack

Go ahead

Ok thank you.
Please reserve L1247 for me.

pinhodecarlos · 2014-04-16, 01:22

About L1247, I got this:

Code:

Tue Apr 15 18:05:09 2014  commencing full merge
Tue Apr 15 18:44:05 2014  memory use: 1649.5 MB
Tue Apr 15 18:44:05 2014  found 91451 cycles, need 7472037
Tue Apr 15 18:44:05 2014  too few cycles, matrix probably cannot build

What do I need to do to keep going? More relations?

frmky · 2014-04-16, 02:40

What matrix density did you specify?

pinhodecarlos · 2014-04-16, 04:32

Quote:

Originally Posted by frmky

What matrix density did you specify?

112.

pinhodecarlos · 2014-04-16, 04:33

Code:

prp80 factor: 44727732200358575718155193552075666016858720751958113346552522635860992435228937
prp136 factor: 4595590077517346285651018529540590684003758485502923471896208062111317688711635008793441657490492808861528525224495090132736838830359533

pinhodecarlos · 2014-04-16, 08:03

L1247 set to target density=90, ETA for LA is 250 hours.

2014-04-14, 15:00	#1289
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 2³·11·73 Posts	On considering cycles to be of equal value I'm being inconsistent with myself. One thing that I've become really sure of, in doing quite large numbers of factorisations with moderately significant resources, is that the circumstances in which it's worth doing more sieving to make the matrix smaller are rare. Particularly rare if you regard watt-hours as the currency to work in, and even more particularly rare if you regard third parties' watt-hours as worth more than yours. For example, for 2+947, I spent a month on 56 threads, 700W of equipment in order to make the linear algebra take two weeks less using 12 threads on 200W equipment. So, if you haven't already pushed more jobs for L1247, don't. It'll take longer even in wall-clock than the time saved, and much longer in terms of coal burned (or, I suppose, water through the Jirau dam). I'll run with the matrix as it is.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Boinc Statistics for NFS@Home borked ?	thomasn	NFS@Home	1	2013-10-02 15:31
BOINC NFS sieving - RSALS	debrouxl	NFS@Home	621	2012-12-14 23:44
BOINC?	masser	Sierpinski/Riesel Base 5	1	2009-02-09 01:10
BOINC?	KEP	Twin Prime Search	212	2007-04-25 10:29
BOINC	bebarce	Software	3	2005-12-15 18:35

2014-04-14, 12:56	#1288
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 2³×11×73 Posts	Carlos: Thanks for your offer, but there's no need. Let the server idle for a while.

2014-04-14, 20:04	#1291
pinhodecarlos "Carlos Pinho" Oct 2011 Milton Keynes, UK 1001101010011₂ Posts	The offer still stands...I can start it tomorrow at this time. Please let me know.

2014-04-14, 22:31	#1292
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 2³×11×73 Posts	Go ahead

2014-04-16, 02:40	#1295
frmky Jul 2003 So Cal 2²·3²·59 Posts	What matrix density did you specify?

2014-04-16, 08:03	#1298
pinhodecarlos "Carlos Pinho" Oct 2011 Milton Keynes, UK 3·17·97 Posts	L1247 set to target density=90, ETA for LA is 250 hours.