Factor not recorded by GPU72

[bayanne]

The following factor does not seem to have been caught by GPUto72:
[Sun May 4 00:36:30 2014]
P-1 found a factor in stage #1, B1=635000.
UID: bayanne/MacPro, M66038201 has a factor: 4201392361479663010747249 (P-1, B1=635000)

[TheJudger]

More details please, I don't see any issues in this report.

Oliver

[ET_]

The factor has more than 81 bits...

[bayanne]

Originally Posted by ET_

The factor has more than 81 bits...

Sorry I don't understand, why?
Is there a previous thread which refers to this?

Previous and subsequent factors have been noted on GPU to 72

What about all these factors which seem to be more than 81 bit
http://www.gpu72.com/reports/largest_factors/

[ET_]

Originally Posted by bayanne

Sorry I don't understand, why?
Is there a previous thread which refers to this?

Previous and subsequent factors have been noted on GPU to 72

What about all these factors which seem to be more than 81 bit
http://www.gpu72.com/reports/largest_factors/

AFAIK, the exponent has been trial-factord only to 73 bits on GPU to 72, so a factor of (more than) 81 bits is welcome but not expected.

Or I missed a part of your question?

Luigi

[bayanne]

Originally Posted by ET_

AFAIK, the exponent has been trial-factord only to 73 bits on GPU to 72, so a factor of (more than) 81 bits is welcome but not expected.

Or I missed a part of your question?

Luigi

The factor appears on PrimeNet but not on GPUto72 stats, just wondering why, which is why I called the thread 'missed factor'. This was a P-1 factor, and other found P-1 factors appear on GPUto72 stats

[chalsall]

Originally Posted by bayanne

The factor appears on PrimeNet but not on GPUto72 stats, just wondering why, which is why I called the thread 'missed factor'. This was a P-1 factor, and other found P-1 factors appear on GPUto72 stats

GPU72 only reports stats on candidates which have been assigned by GPU72 for TF'ing or P-1'ing. This candidate was assigned for TF'ing (from 71 to 73, to user "Aaron Haviland"), but then returned to Primenet.

Presumably you got the assignment from Primenet sometime later, and then found the factor.

[bayanne]

Originally Posted by chalsall

GPU72 only reports stats on candidates which have been assigned by GPU72 for TF'ing or P-1'ing. This candidate was assigned for TF'ing (from 71 to 73, to user "Aaron Haviland"), but then returned to Primenet.

Presumably you got the assignment from Primenet sometime later, and then found the factor.

I received the exponent via the gimps.gpu72.com proxy which I have been using for somewhile. I did not receive it direct from PrimeNet

Still don't understand ...

[kracker]

Originally Posted by bayanne

I received the exponent via the gimps.gpu72.com proxy which I have been using for somewhile. I did not receive it direct from PrimeNet

Still don't understand ...

I don't think we understand either. You're saying missed factor, how? Seems normal.

[chalsall]

Originally Posted by bayanne

I received the exponent via the gimps.gpu72.com proxy which I have been using for somewhile. I did not receive it direct from PrimeNet

Still don't understand ...

Nor do I...

OK, I will drill down on the logs, and see what I can see. But according to the database this was not assigned to you by GPU72 -- either manually or via the Proxy. But the raw logs might tell me more.

If it helps at all, I will manually create an assignment record for this work in the interim, so you'll see the credits on GPU72 (probably not until tomorrow).

[chalsall]

Originally Posted by chalsall

OK, I will drill down on the logs, and see what I can see.

OK... I now understand what happened here. And thank you -- this points out a subtle bug with the proxy...

Your machine asked for work type 101 (Double Check), but GPU72 no longer offers this because of the new Primenet assignment rules, so it immediately deferred the request to Primenet for fulfillment.

Rather than Primenet giving your machine DC work, it instead gave you P-1 work (I can't tell you why -- perhaps your memory settings, or perhaps your preference settings).

If Primenet had given you DC work GPU72 would have tracked this. As it gave you P-1 work instead it didn't.

As I said above, I will create a manual assignment record so the system picks up this work and this factor. And, sometime (probably in the distant future) I might have time to fix this rare bug.

GetAssignment -- 8ef3....29a8 (3) wants WorkType 101

DeferToPrimenet:
Response:
pnErrorResult=0
pnErrorDetail=Server assigned P-1 factoring work.
g=8ef3....29a8
k=F825....C3C0
A=1
b=2
n=66038201
c=-1
w=4
sf=73
saved=2
==END==