[QUOTE=chalsall;303654]OK. I've changed the report to show "Completed" in the Total column for those ranges where the work has been, well, completed....[/QUOTE]

Then, call me slow but I am still a little confused.

What then does the 20,808 count represent; I thought it might be exponents left to factor but apparently it is not. It is also NOT total exponents ...AHHHHH...now I know;:fusion: it is the exponents left to DC to the desired bit level...DING DING DING!!!!

[QUOTE=petrw1;303664]...AHHHHH...now I know;:fusion: it is the exponents left to DC to the desired bit level...DING DING DING!!!![/QUOTE]

Correct.

I have added some language at the bottom of the page to hopefully explain what the report shows more clearly.

Due to the heat here we had to clock back our boxes to 2GHz each. Fortunately that is still enough to saturate a GTX 570 with four instances running with SievePrimes at 5000. (Three instances would actually work, but we run four to get that last 1-2% and also in case one worker dies for some reason.)

Surprisingly, this does not cut our production as much as it would sound. It sure makes a difference in room temperature though!

957,286.1 GHzd [URL="http://www.gpu72.com/reports/factoring_cost/"]per factor[/URL] does seem a little off, doesn't it? :mike:

(LLTF, average to 2^72 ... actual number fluctuates, of course.)

[QUOTE=ckdo;303738]957,286.1 GHzd [URL="http://www.gpu72.com/reports/factoring_cost/"]per factor[/URL] does seem a little off, doesn't it? :mike:[/QUOTE]

Sigh... Thanks. Fixed.

Looks like we are a couple days away from taking the #2 team spot, and a couple weeks from hitting 1,000,000 GHzD (would that make it a PHzD?)

[QUOTE=bcp19;303915]Looks like we are a couple days away from taking the #2 team spot, and a couple weeks from hitting 1,000,000 GHzD (would that make it a PHzD?)[/QUOTE]

We moved into the [URL="http://www.mersenne.org/report_top_teams/"]#2 spot (for the last year)[/URL]sometime today.

Based on a quick calculation, we should pass 1M GHzDs in less than a week!

Thanks for all the cycles everyone!!! :smile:

[QUOTE=chalsall;303988]Based on a quick calculation, we should pass 1M GHzDs in less than a week![/QUOTE]

...saved...

Missing completed assignments
 

Hi all!

I think that I have found a possible problem in the code that collects the result. I'm trying to understand what the problem is...

On [SIZE=2]2012-07-04 21:28 I reported 15 lines of results to primenet. One factor and 14 negative results. Out of these only four was registered in GPU to 72.[/SIZE]

[SIZE=2]62540461 [/SIZE]factor [SIZE=2]898550098924501937929[/SIZE] not registered in GPU to 72
[SIZE=2]54569219[/SIZE][SIZE=2] Complete result and correctly registered in GPU to 72[/SIZE]
[SIZE=2]54567179[/SIZE][SIZE=2] Complete result and correctly registered in GPU to 72[/SIZE]
[SIZE=2]26881103[/SIZE][SIZE=2] Complete result and correctly registered in GPU to 72[/SIZE]
[SIZE=2]63285907[/SIZE][SIZE=2] Complete result but not registered in GPU to 72[/SIZE]
[SIZE=2]62130143[/SIZE][SIZE=2] Complete result but not registered in GPU to 72[/SIZE]
[SIZE=2]61668697[/SIZE][SIZE=2] Complete result but not registered in GPU to 72[/SIZE]
[SIZE=2]61668689[/SIZE][SIZE=2] Complete result but not registered in GPU to 72[/SIZE]
[SIZE=2]61666729[/SIZE][SIZE=2] Complete result but not registered in GPU to 72[/SIZE]
[SIZE=2]61666141 [/SIZE][SIZE=2]Complete result but not registered in GPU to 72[/SIZE]
[SIZE=2]61466891 Complete result but not registered in GPU to 72
[/SIZE][SIZE=2]61429351 Partial result, not finished yet
[/SIZE][SIZE=2]61428923 [/SIZE][SIZE=2]Partial result, full credit given when later result came in.[/SIZE]
[SIZE=2]46313123 Partial result, full credit given when later result came in.
[/SIZE][SIZE=2]46282337 Partial result reported earlier. Full credit given when this result came in.

Missed factors later dates...
[/SIZE][SIZE=2]2012-07-04 22:34[/SIZE][SIZE=2]Complete assignment but not registered in GPU to 72
[/SIZE][SIZE=2]63343723[/SIZE]
[SIZE=2]2012-07-05 08:18[/SIZE][SIZE=2]Complete assignment but not registered in GPU to 72
[/SIZE][SIZE=2]63343727[/SIZE] [SIZE=2]63343793 [/SIZE][SIZE=2]63343927 [/SIZE][SIZE=2]63343949[/SIZE] [SIZE=2]63343997 [/SIZE][SIZE=2]63344041 [/SIZE][SIZE=2]63344077[/SIZE]

Ok, I think I see the pattern here... the code was limiting the reports of finished candidates. Strangely the candidates were cleared from my assignment.
I guess the problem have already been fixed as I got M[SIZE=2]62050271 in my Completed report and that is larger than [/SIZE]M[SIZE=2]61668697[/SIZE] that is missing...

As this is my first post I just want to add that this is a great project, it gives some purpose to my GPU (HD 6970). :smile: :tu:

/Göran

[QUOTE=Axelsson;304228]Ok, I think I see the pattern here... the code was limiting the reports of finished candidates. Strangely the candidates were cleared from my assignment.

I guess the problem have already been fixed as I got M[SIZE=2]62050271 in my Completed report and that is larger than [/SIZE]M[SIZE=2]61668697[/SIZE] that is missing...[/QUOTE]

Nope... The "problem" was simply that GPU72 doesn't recognize any completions which it itself didn't assign. It appears from your list of candidates above that you've been doing some from GPU72, and some directly from PrimeNet. That's fine, of course -- PrimeNet is canonical on the state of each candidate -- but GPU72 won't award credit to you for work it didn't assign to you, even if you did it.

It would put too much load on PrimeNet for the system to determine who completed what assignments which "Spidy" doesn't know to watch out for. (It's actually a little bit more complicated than that, but the technical details are immaterial to what you've observed.)

[QUOTE=Axelsson;304228]As this is my first post I just want to add that this is a great project, it gives some purpose to my GPU (HD 6970). :smile: :tu:[/QUOTE]

Thanks very much. And thanks for participating. :smile:

Strange... I have to my knowledge only reserved one candidate directly from primenet and that was more than a month before just as a test. I think I got those candidates at 61M - 63M from GPU72 but I might be mistaken... I've been wrong before but that was a mistake. :grin:

/Göran