GPU to 72 status... - Page 114

chalsall · 2012-05-23, 18:22

Quote:

Originally Posted by petrw1

Not yet....still nothing below 28M and most at 29M

Yeah... That's as expected.

Spidy doesn't have any special privilages from PrimeNet. The queue will get lower exponents as they become available early tomorrow morning (UTC).

petrw1 · 2012-05-28, 04:02

http://www.gpu72.com/reports/overall/graph/month/

Or is this just a sign of too much progress (i.e. caught up) and less critical of a need?

chalsall · 2012-05-28, 14:42

Quote:

Originally Posted by petrw1

Or is this just a sign of too much progress (i.e. caught up) and less critical of a need?

We're pretty steady-state at the moment; and have been for the last three months.

Don't forget that our #2 contributor, Xyzzy, only submits once a week or so. That explains the lower "GHz Days per Day of Work Done" showing on the graph over the last few days.

However, with regards to "caught up", that's definitely the case in the DC range. We are currently over 400 days ahead of that wave. And, while I'm "talking", I had might as well mention that a few days ago the last few DC candidates at 67 bits (in the 36M and 37M ranges) were brought up to 68.

nucleon · 2012-05-29, 00:22

Work saved is dropping as we're doing higher bit levels now. (higher bit levels = more GHz-days needed to yield a factor; work saved = f(factors found, etc) ) To be expected.

Is Xyzzy submitting all results now? (not just factors found)

I thought that might be cause for slowness trend.

-- Craig Meyers

chalsall · 2012-05-29, 01:07

Quote:

Originally Posted by nucleon

Work saved is dropping as we're doing higher bit levels now. (higher bit levels = more GHz-days needed to yield a factor; work saved = f(factors found, etc) ) To be expected.

Good point. While we're working in the higher ranges, where each factor is worth more using the "GHz Days Saved" metric, we're also effectively only working from 71->72 (or ->73) which takes more work per factor found, as all the "low hanging fruit" has already been harvested at the lower bit levels.

Quote:

Originally Posted by nucleon

Is Xyzzy submitting all results now? (not just factors found)

Yes.

Xyzzy · 2012-05-29, 01:36

Quote:

Is Xyzzy submitting all results now?

Since George tweaked PrimeNet it will now accept 3000+ line uploads.

James Heinrich · 2012-05-29, 02:24

Quote:

Originally Posted by Xyzzy

Since George tweaked PrimeNet it will now accept 3000+ line uploads.

I have noticed this too... I just uploaded 2501 lines myself.
Of course, my 2k5 lines (about 2 weeks work) totaled 368GHz-days... or about 5 nucleon-hours

flashjh · 2012-05-29, 03:02

Quote:

Originally Posted by James Heinrich

about 5 nucleon-hours

Graff · 2012-05-29, 14:43

Last night I was down to 27 pending P-1 assignments. So I did my usual
weekly grab of 80 exponents (about a week's work). I only got 70
exponents. So I grabbed another ten. 80 exponents, distributed around
my various machines.

Check the P-1 factoring progress report, see that only 77 exponents are
assigned to me. Now, this morning, it's down to 73. That's only 50
exponents assigned. What happened to the other 30 that I see in the
file I downloaded from GPUto72?

Gareth

chalsall · 2012-05-29, 15:02

Quote:

Originally Posted by Graff

Check the P-1 factoring progress report, see that only 77 exponents are assigned to me. Now, this morning, it's down to 73. That's only 50
exponents assigned. What happened to the other 30 that I see in the
file I downloaded from GPUto72?

Not sure.

Could you please PM me a couple of examples of P-1 assignments which you don't see any more?

Graff · 2012-05-29, 16:31

Quote:

Originally Posted by chalsall

Not sure.

Could you please PM me a couple of examples of P-1 assignments which you don't see any more?

Sent via PM.

Gareth

2012-05-28, 04:02	#1245
petrw1 1976 Toyota Corona years forever! "Wayne" Nov 2006 Saskatchewan, Canada 3×5×313 Posts	Are we slowing? http://www.gpu72.com/reports/overall/graph/month/ Or is this just a sign of too much progress (i.e. caught up) and less critical of a need?

2012-05-29, 14:43	#1252
Graff Jul 2006 USA (UT-5) via UK (UT) 11101100₂ Posts	Missing P-1 assignments? Last night I was down to 27 pending P-1 assignments. So I did my usual weekly grab of 80 exponents (about a week's work). I only got 70 exponents. So I grabbed another ten. 80 exponents, distributed around my various machines. Check the P-1 factoring progress report, see that only 77 exponents are assigned to me. Now, this morning, it's down to 73. That's only 50 exponents assigned. What happened to the other 30 that I see in the file I downloaded from GPUto72? Gareth

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Status	Primeinator	Operation Billion Digits	5	2011-12-06 02:35
62 bit status	1997rj7	Lone Mersenne Hunters	27	2008-09-29 13:52
OBD Status	Uncwilly	Operation Billion Digits	22	2005-10-25 14:05
1-2M LLR status	paulunderwood	3*2^n-1 Search	2	2005-03-13 17:03
Status of 26.0M - 26.5M	1997rj7	Lone Mersenne Hunters	25	2004-06-18 16:46

2012-05-29, 00:22	#1247
nucleon Mar 2003 Melbourne 5·103 Posts	Work saved is dropping as we're doing higher bit levels now. (higher bit levels = more GHz-days needed to yield a factor; work saved = f(factors found, etc) ) To be expected. Is Xyzzy submitting all results now? (not just factors found) I thought that might be cause for slowness trend. -- Craig Meyers